Skip to content

SeetaResearch / Dragon

Go to a project
Commit b6182abc by Ting PAN
Options

add examples/Seg-FCN

1 parent 98b31daf
Inline Side-by-side
Showing with 8402 additions and 0 deletions
examples/README.md
......@@ -10,4 +10,6 @@ which was described in our arXiv paper: [Dragon: A Computation Graph Virtual Mac
* [cifar10](https://github.com/neopenx/Dragon/tree/master/examples/cifar10) - How to train/infer a basic classification network [*Caffe1 Style*]
* [Seg-FCN](https://github.com/neopenx/Dragon/tree/master/examples/Seg-FCN) - Fully Convolutional Networks for Semantic Segmentation [*Caff1 Style*]
* [GA3C](https://github.com/neopenx/Dragon/tree/master/examples/GA3C) - A hybrid CPU/GPU version of the A3C algorithm [*TinyDragon Style*]
examples/Seg-FCN/README.md 0 → 100644
# Fully Convolutional Networks for Semantic Segmentation
This is the reference implementation of the models and code for the fully convolutional networks (FCNs) in the [PAMI FCN](https://arxiv.org/abs/1605.06211) and [CVPR FCN](http://www.cv-foundation.org/openaccess/content_cvpr_2015/html/Long_Fully_Convolutional_Networks_2015_CVPR_paper.html) papers:
Fully Convolutional Models for Semantic Segmentation
Evan Shelhamer*, Jonathan Long*, Trevor Darrell
PAMI 2016
arXiv:1605.06211
Fully Convolutional Models for Semantic Segmentation
Jonathan Long*, Evan Shelhamer*, Trevor Darrell
CVPR 2015
arXiv:1411.4038
### Requirements: software
Python packages you might not have: `numpy`, `PIL`, `python-opencv`
### Requirements: hardware
For training the FCN with VGG16 for VOC images(~500x350), 4G of GPU memory is sufficient (using CUDNN)
### Installation (sufficient for the demo)
1. We'll call the directory of Seg-FCN as `FCN_ROOT`
2. Download pre-computed Seg-FCN models
* [FCN-32s PASCAL](http://dl.caffe.berkeleyvision.org/fcn32s-heavy-pascal.caffemodel): single stream, 32 pixel prediction stride net, scoring 63.6 mIU on seg11valid
* [FCN-16s PASCAL](http://dl.caffe.berkeleyvision.org/fcn16s-heavy-pascal.caffemodel): two stream, 16 pixel prediction stride net, scoring 65.0 mIU on seg11valid
* [FCN-8s PASCAL](http://dl.caffe.berkeleyvision.org/fcn8s-heavy-pascal.caffemodel): three stream, 8 pixel prediction stride net, scoring 65.5 mIU on seg11valid and 67.2 mIU on seg12test
* [FCN-8s PASCAL at-once](http://dl.caffe.berkeleyvision.org/fcn8s-atonce-pascal.caffemodel): all-at-once, three stream, 8 pixel prediction stride net, scoring 65.4 mIU on seg11valid
```Shell
cp fcn8s-heavy-pascal.caffemodel $FCN_ROOT/data/seg_fcn_models
```
These models were trained online with high momentum, using extra data from [Hariharan et al.](http://www.cs.berkeley.edu/~bharath2/codes/SBD/download.html), but excluding SBD val.
FCN-32s is fine-tuned from the [ILSVRC-trained VGG-16 model](https://github.com/BVLC/caffe/wiki/Model-Zoo#models-used-by-the-vgg-team-in-ilsvrc-2014), and the finer strides are then fine-tuned in turn.
The "at-once" FCN-8s is fine-tuned from VGG-16 all-at-once by scaling the skip connections to better condition optimization.
### Demo
*After successfully completing [basic installation](#installation-sufficient-for-the-demo)*, you'll be ready to run the demo.
To run the demo
```Shell
cd $FCN_ROOT
python infer.py
```
The demo performs semantic segmentation using a VGG16 network trained for semantic segmentation on SBDD.
### Beyond the demo: installation for training and testing models
1. Download the SBDD(for training), VOC2011(for testing)
```Shell
wget http://www.eecs.berkeley.edu/Research/Projects/CS/vision/grouping/semantic_contours/benchmark.tgz
wget http://host.robots.ox.ac.uk/pascal/VOC/voc2011/VOCtrainval_25-May-2011.tar
```
2. Extract all of these tars into the $FCN_ROOT/data, should have this basic structure
```Shell
$FCN_ROOT/data/sbdd/dataset
$FCN_ROOT/data/pascal/VOC2011
# ... and several other directories ...
```
3. Follow the next sections to download pre-trained ImageNet models
### Download pre-trained ImageNet models
Pre-trained [ImageNet models](http://pan.baidu.com/s/1eSGLwsE) can be downloaded for backbone net: VGG16.
### Transplant a fully-connected net into a fully-convolution net
```Shell
cp VGG16.v2.caffemodel $FCN_ROOT/transplant/VGG16
cd $FCN_ROOT/transplant/VGG16
python solve.py
```
This script will generate a new model ``VGG16.fcn.caffemodel`` for training.
### Training
FCN prefers two training methods:
1. CVPR version:
First, Train FCN-32s for 1 day.
Then, Train FCN-16s fintune from FCN-32s for 1 day.
Final, Train FCN-8s fintune from FCN-16s for 1 day.
Follow this way, you should run $FCN_ROOT/voc-fcn32s | voc-fcn16s | fcn-8s/solve.py ``sequentially``.
2. PAMI version:
Directly run $FCN_ROOT/voc-fcn8s-atonce/solve.py
Both of above ways train same iterations, ``PAMI ver.`` is simpier and got similar results.
Trained Seg-FCN networks are saved under:
```
voc-fcnxs/snapshot/
```
Test outputs are saved under:
```
voc-fcnxs/segs/
```
examples/Seg-FCN/colors/pascal_voc.act 0 → 100644
No preview for this file type
examples/Seg-FCN/data/seg11valid.txt 0 → 100644
2007_000033
2007_000042
2007_000061
2007_000123
2007_000129
2007_000175
2007_000187
2007_000323
2007_000332
2007_000346
2007_000452
2007_000464
2007_000491
2007_000529
2007_000559
2007_000572
2007_000629
2007_000636
2007_000661
2007_000663
2007_000676
2007_000727
2007_000762
2007_000783
2007_000799
2007_000804
2007_000830
2007_000837
2007_000847
2007_000862
2007_000925
2007_000999
2007_001154
2007_001175
2007_001239
2007_001284
2007_001288
2007_001289
2007_001299
2007_001311
2007_001321
2007_001377
2007_001408
2007_001423
2007_001430
2007_001457
2007_001458
2007_001526
2007_001568
2007_001585
2007_001586
2007_001587
2007_001594
2007_001630
2007_001677
2007_001678
2007_001717
2007_001733
2007_001761
2007_001763
2007_001774
2007_001884
2007_001955
2007_002046
2007_002094
2007_002119
2007_002132
2007_002260
2007_002266
2007_002268
2007_002284
2007_002376
2007_002378
2007_002387
2007_002400
2007_002412
2007_002426
2007_002427
2007_002445
2007_002470
2007_002539
2007_002565
2007_002597
2007_002618
2007_002619
2007_002624
2007_002643
2007_002648
2007_002719
2007_002728
2007_002823
2007_002824
2007_002852
2007_002903
2007_003011
2007_003020
2007_003022
2007_003051
2007_003088
2007_003101
2007_003106
2007_003110
2007_003131
2007_003134
2007_003137
2007_003143
2007_003169
2007_003188
2007_003194
2007_003195
2007_003201
2007_003349
2007_003367
2007_003373
2007_003499
2007_003503
2007_003506
2007_003530
2007_003571
2007_003587
2007_003611
2007_003621
2007_003682
2007_003711
2007_003714
2007_003742
2007_003786
2007_003841
2007_003848
2007_003861
2007_003872
2007_003917
2007_003957
2007_003991
2007_004033
2007_004052
2007_004112
2007_004121
2007_004143
2007_004189
2007_004190
2007_004193
2007_004241
2007_004275
2007_004281
2007_004380
2007_004392
2007_004405
2007_004468
2007_004483
2007_004510
2007_004538
2007_004558
2007_004644
2007_004649
2007_004712
2007_004722
2007_004856
2007_004866
2007_004902
2007_004969
2007_005058
2007_005074
2007_005107
2007_005114
2007_005149
2007_005173
2007_005281
2007_005294
2007_005296
2007_005304
2007_005331
2007_005354
2007_005358
2007_005428
2007_005460
2007_005469
2007_005509
2007_005547
2007_005600
2007_005608
2007_005626
2007_005689
2007_005696
2007_005705
2007_005759
2007_005803
2007_005813
2007_005828
2007_005844
2007_005845
2007_005857
2007_005911
2007_005915
2007_005978
2007_006028
2007_006035
2007_006046
2007_006076
2007_006086
2007_006117
2007_006171
2007_006241
2007_006260
2007_006277
2007_006348
2007_006364
2007_006373
2007_006444
2007_006449
2007_006549
2007_006553
2007_006560
2007_006647
2007_006678
2007_006680
2007_006698
2007_006761
2007_006802
2007_006837
2007_006841
2007_006864
2007_006866
2007_006946
2007_007007
2007_007084
2007_007109
2007_007130
2007_007165
2007_007168
2007_007195
2007_007196
2007_007203
2007_007211
2007_007235
2007_007341
2007_007414
2007_007417
2007_007470
2007_007477
2007_007493
2007_007498
2007_007524
2007_007534
2007_007624
2007_007651
2007_007688
2007_007748
2007_007795
2007_007810
2007_007815
2007_007818
2007_007836
2007_007849
2007_007881
2007_007996
2007_008051
2007_008084
2007_008106
2007_008110
2007_008204
2007_008222
2007_008256
2007_008260
2007_008339
2007_008374
2007_008415
2007_008430
2007_008543
2007_008547
2007_008596
2007_008645
2007_008670
2007_008708
2007_008722
2007_008747
2007_008802
2007_008815
2007_008897
2007_008944
2007_008964
2007_008973
2007_008980
2007_009015
2007_009068
2007_009084
2007_009088
2007_009096
2007_009221
2007_009245
2007_009251
2007_009252
2007_009258
2007_009320
2007_009323
2007_009331
2007_009346
2007_009392
2007_009413
2007_009419
2007_009446
2007_009458
2007_009521
2007_009562
2007_009592
2007_009654
2007_009655
2007_009684
2007_009687
2007_009691
2007_009706
2007_009750
2007_009756
2007_009764
2007_009794
2007_009817
2007_009841
2007_009897
2007_009911
2007_009923
2007_009938
2008_000073
2008_000075
2008_000107
2008_000123
2008_000149
2008_000213
2008_000215
2008_000223
2008_000233
2008_000239
2008_000271
2008_000345
2008_000391
2008_000401
2008_000501
2008_000533
2008_000573
2008_000589
2008_000657
2008_000661
2008_000725
2008_000731
2008_000763
2008_000765
2008_000811
2008_000853
2008_000911
2008_000919
2008_000943
2008_001135
2008_001231
2008_001249
2008_001379
2008_001433
2008_001439
2008_001513
2008_001531
2008_001547
2008_001715
2008_001821
2008_001885
2008_001971
2008_002043
2008_002205
2008_002239
2008_002269
2008_002273
2008_002379
2008_002383
2008_002467
2008_002521
2008_002623
2008_002681
2008_002775
2008_002835
2008_002859
2008_003105
2008_003135
2008_003155
2008_003369
2008_003709
2008_003777
2008_003821
2008_003885
2008_004069
2008_004172
2008_004175
2008_004279
2008_004339
2008_004345
2008_004363
2008_004453
2008_004562
2008_004575
2008_004621
2008_004659
2008_004705
2008_004995
2008_005049
2008_005097
2008_005105
2008_005145
2008_005217
2008_005262
2008_005439
2008_005525
2008_005633
2008_005637
2008_005691
2008_006055
2008_006229
2008_006327
2008_006553
2008_006835
2008_007025
2008_007031
2008_007123
2008_007497
2008_007677
2008_007797
2008_007811
2008_008051
2008_008103
2008_008301
2009_000013
2009_000022
2009_000032
2009_000037
2009_000039
2009_000087
2009_000121
2009_000149
2009_000201
2009_000205
2009_000219
2009_000335
2009_000351
2009_000387
2009_000391
2009_000446
2009_000455
2009_000457
2009_000469
2009_000487
2009_000523
2009_000619
2009_000641
2009_000675
2009_000705
2009_000723
2009_000727
2009_000771
2009_000845
2009_000879
2009_000919
2009_000931
2009_000935
2009_000989
2009_000991
2009_001255
2009_001299
2009_001333
2009_001363
2009_001391
2009_001411
2009_001433
2009_001505
2009_001535
2009_001565
2009_001607
2009_001663
2009_001683
2009_001687
2009_001731
2009_001775
2009_001851
2009_001941
2009_002035
2009_002165
2009_002171
2009_002221
2009_002291
2009_002295
2009_002317
2009_002445
2009_002487
2009_002521
2009_002527
2009_002535
2009_002539
2009_002549
2009_002571
2009_002573
2009_002591
2009_002635
2009_002649
2009_002651
2009_002727
2009_002749
2009_002753
2009_002771
2009_002887
2009_002975
2009_003003
2009_003005
2009_003059
2009_003063
2009_003065
2009_003071
2009_003105
2009_003123
2009_003193
2009_003269
2009_003273
2009_003311
2009_003323
2009_003343
2009_003387
2009_003481
2009_003517
2009_003523
2009_003549
2009_003551
2009_003589
2009_003607
2009_003703
2009_003707
2009_003771
2009_003849
2009_003857
2009_003895
2009_004021
2009_004033
2009_004043
2009_004099
2009_004125
2009_004217
2009_004255
2009_004455
2009_004507
2009_004509
2009_004579
2009_004581
2009_004687
2009_004801
2009_004859
2009_004867
2009_004895
2009_004969
2009_004993
2009_005087
2009_005089
2009_005137
2009_005189
2009_005217
2009_005219
2010_000003
2010_000065
2010_000083
2010_000159
2010_000163
2010_000309
2010_000427
2010_000559
2010_000573
2010_000639
2010_000683
2010_000907
2010_000961
2010_001017
2010_001061
2010_001069
2010_001149
2010_001151
2010_001251
2010_001313
2010_001327
2010_001331
2010_001553
2010_001557
2010_001563
2010_001577
2010_001579
2010_001767
2010_001773
2010_001851
2010_001995
2010_002017
2010_002025
2010_002137
2010_002147
2010_002161
2010_002271
2010_002305
2010_002361
2010_002531
2010_002623
2010_002693
2010_002701
2010_002763
2010_002921
2010_002929
2010_002939
2010_003123
2010_003187
2010_003207
2010_003239
2010_003275
2010_003325
2010_003365
2010_003381
2010_003409
2010_003453
2010_003473
2010_003495
2010_003531
2010_003547
2010_003675
2010_003781
2010_003813
2010_003915
2010_003971
2010_004041
2010_004063
2010_004149
2010_004165
2010_004219
2010_004355
2010_004419
2010_004479
2010_004529
2010_004543
2010_004551
2010_004559
2010_004697
2010_004763
2010_004783
2010_004795
2010_004815
2010_004825
2010_005013
2010_005021
2010_005063
2010_005159
2010_005187
2010_005245
2010_005305
2010_005421
2010_005531
2010_005705
2010_005709
2010_005719
2010_005727
2010_005871
2010_005877
2010_005899
2010_005991
2011_000045
2011_000051
2011_000173
2011_000185
2011_000291
2011_000419
2011_000435
2011_000455
2011_000479
2011_000503
2011_000521
2011_000536
2011_000598
2011_000607
2011_000661
2011_000669
2011_000747
2011_000789
2011_000809
2011_000843
2011_000969
2011_001069
2011_001071
2011_001161
2011_001263
2011_001281
2011_001287
2011_001313
2011_001341
2011_001421
2011_001447
2011_001529
2011_001567
2011_001589
2011_001597
2011_001601
2011_001607
2011_001613
2011_001619
2011_001665
2011_001669
2011_001713
2011_001745
2011_001775
2011_001793
2011_001812
2011_001868
2011_001984
2011_002041
2011_002121
2011_002223
2011_002279
2011_002295
2011_002317
2011_002327
2011_002343
2011_002371
2011_002379
2011_002391
2011_002509
2011_002535
2011_002575
2011_002589
2011_002623
2011_002641
2011_002675
2011_002685
2011_002713
2011_002863
2011_002929
2011_002993
2011_002997
2011_003011
2011_003055
2011_003085
2011_003145
2011_003197
2011_003271
\ No newline at end of file
examples/Seg-FCN/infer.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Source Code by Evan Shelhamer
# Re-Written by Ting Pan
# --------------------------------------------------------
""" Infer for a single Image and show """
import numpy as np
from PIL import Image
import dragon.vm.caffe as caffe
import dragon.core.workspace as ws
import os
import cv2
# init
caffe.set_mode_gpu()
# load net
net = caffe.Net('voc-fcn8s/deploy.prototxt', 'data/seg_fcn_models/fcn8s-heavy-pascal.caffemodel', caffe.TEST)
# load color table
color_table = np.fromfile('colors/pascal_voc.act', dtype=np.uint8)
def load_image(file):
# load image, switch to BGR, subtract mean, and make dims C x H x W for Caffe
im = Image.open(file)
in_ = np.array(im, dtype=np.float32)
in_ = in_[:,:,::-1]
in_ -= np.array((104.00698793,116.66876762,122.67891434))
in_ = in_.transpose((2,0,1))
return in_
def seg(file, save_dir="data/seg_results", mix=True, show=True):
if save_dir is not None:
if not os.path.exists(save_dir):
os.makedirs(save_dir)
im = load_image(file)
# shape for input (data blob is N x C x H x W), set data
im = im.reshape(1, *im.shape)
ws.FeedTensor(net.blobs['data'].data, im)
# run net and take argmax for prediction
net.forward()
if save_dir is not None:
filename_ext = file.split('/')[-1]
filename = filename_ext.split('.')[-2]
filepath = os.path.join(save_dir, filename + '.png')
mat = ws.FetchTensor(net.blobs['score'].data)
im = Image.fromarray(mat[0].argmax(0).astype(np.uint8), mode='P')
im.putpalette(color_table)
im.save(filepath)
if show:
if mix:
show1 = cv2.imread(file)
show2 = cv2.imread(filepath)
show3 = cv2.addWeighted(show1, 0.7, show2, 0.5, 1)
else: show3 = cv2.imread(filepath)
cv2.imshow('Seg-FCN', show3)
cv2.waitKey(0)
if __name__ == '__main__':
seg('data/demo/001763.jpg')
examples/Seg-FCN/score.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Source Code by Evan Shelhamer
# Re-Written by Ting Pan
# --------------------------------------------------------
from __future__ import division
import dragon.core.workspace as ws
import numpy as np
import os
from datetime import datetime
from PIL import Image
color_table = np.fromfile('../colors/pascal_voc.act', dtype=np.uint8)
def fast_hist(a, b, n):
k = (a >= 0) & (a < n)
return np.bincount(n * a[k].astype(int) + b[k], minlength=n**2).reshape(n, n)
def compute_hist(net, save_dir, dataset, layer='score', gt='label'):
n_cl = hist = None
loss = 0
for idx in dataset:
net.forward()
gt_mat = ws.FetchTensor(net.blobs[gt].data)
layer_mat = ws.FetchTensor(net.blobs[layer].data)
loss_mat = ws.FetchTensor(net.blobs['loss'].data)
if n_cl is None: n_cl = layer_mat.shape[1]
if hist is None: hist = np.zeros((n_cl, n_cl))
hist += fast_hist(gt_mat[0, 0].flatten(),
layer_mat[0].argmax(0).flatten(), n_cl)
if save_dir:
im = Image.fromarray(layer_mat[0].argmax(0).astype(np.uint8), mode='P')
im.putpalette(color_table)
im.save(os.path.join(save_dir, idx + '.png'))
# compute the loss as well
loss += loss_mat.flat[0]
return hist, loss / len(dataset)
def seg_tests(solver, save_format, dataset, layer='score', gt='label'):
print '>>>', datetime.now(), 'Begin seg tests'
solver.test_nets[0].share_with(solver.net)
do_seg_tests(solver.test_nets[0], solver.iter, save_format, dataset, layer, gt)
def do_seg_tests(net, iter, save_format, dataset, layer='score', gt='label'):
if save_format:
save_format = save_format.format(iter)
if not os.path.exists(save_format): os.makedirs(save_format)
hist, loss = compute_hist(net, save_format, dataset, layer, gt)
# mean loss
print '>>>', datetime.now(), 'Iteration', iter, 'loss', loss
# overall accuracy
acc = np.diag(hist).sum() / hist.sum()
print '>>>', datetime.now(), 'Iteration', iter, 'overall accuracy', acc
# per-class accuracy
acc = np.diag(hist) / hist.sum(1)
print '>>>', datetime.now(), 'Iteration', iter, 'mean accuracy', np.nanmean(acc)
# per-class IU
iu = np.diag(hist) / (hist.sum(1) + hist.sum(0) - np.diag(hist))
print '>>>', datetime.now(), 'Iteration', iter, 'mean IU', np.nanmean(iu)
freq = hist.sum(1) / hist.sum()
print '>>>', datetime.now(), 'Iteration', iter, 'fwavacc', \
(freq[freq > 0] * iu[freq > 0]).sum()
return hist
examples/Seg-FCN/surgery.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Source Code by Evan Shelhamer
# Re-Written by Ting Pan
# --------------------------------------------------------
from __future__ import division
import dragon.core.workspace as ws
import numpy as np
def transplant(new_net, net):
func = net.function; func = new_net.function
for p in net.params:
if p not in new_net.params:
print 'dropping', p
continue
for i in range(len(net.params[p])):
if i > (len(new_net.params[p]) - 1):
print 'dropping', p, i
break
print 'copying', p, i
net_param = ws.FetchTensor(net.params[p][i].data)
new_net_param = ws.FetchTensor(new_net.params[p][i].data)
name = new_net.params[p][i].data._name
if net_param.shape != new_net_param.shape:
print 'coercing', p, i, 'from', net_param.shape, 'to', new_net_param.shape
else:
pass
new_net_param.flat = new_net_param.flat
ws.FeedTensor(name, new_net_param)
def upsample_filt(size):
factor = (size + 1) // 2
if size % 2 == 1:
center = factor - 1
else:
center = factor - 0.5
og = np.ogrid[:size, :size]
return (1 - abs(og[0] - center) / factor) * \
(1 - abs(og[1] - center) / factor)
def interp(net, layers):
print 'bilinear-interp for layers:', layers
net.forward() # dragon must forward once to create weights
for l in layers:
net_param = ws.FetchTensor(net.params[l][0].data)
m, k, h, w = net_param.shape
if m != k and k != 1:
print 'input + output channels need to be the same or |output| == 1'
raise
if h != w:
print 'filters need to be square'
raise
filt = upsample_filt(h)
net_param[range(m), range(k), :, :] = filt
ws.FeedTensor(net.params[l][0].data._name, net_param)
examples/Seg-FCN/transplants/VGG16/net.prototxt 0 → 100644
input: "data"
input_shape {
dim: 1
dim: 3
dim: 224
dim: 224
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "InnerProduct"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
inner_product_param {
num_output: 4096
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
\ No newline at end of file
examples/Seg-FCN/transplants/VGG16/new_net.prototxt 0 → 100644
input: "data"
input_shape {
dim: 1
dim: 3
dim: 224
dim: 224
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
\ No newline at end of file
examples/Seg-FCN/transplants/VGG16/solve.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Transplant fully-connected caffemodel into fully-convolution ver. """
import surgery
import dragon.vm.caffe as caffe
if __name__ == '__main__':
net = caffe.Net('net.prototxt', 'VGG16.v2.caffemodel', caffe.TEST)
new_net = caffe.Net('new_net.prototxt', caffe.TEST)
surgery.transplant(new_net, net)
new_net.save('VGG16.fcn.caffemodel', suffix='')
\ No newline at end of file
examples/Seg-FCN/voc-fcn16s/caffemodel-url 0 → 100644
http://dl.caffe.berkeleyvision.org/fcn16s-heavy-pascal.caffemodel
\ No newline at end of file
examples/Seg-FCN/voc-fcn16s/net.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Source Code by Evan Shelhamer
# Re-Written by Ting Pan
# --------------------------------------------------------
import dragon.vm.caffe as caffe
from dragon.vm.caffe import layers as L, params as P
from dragon.vm.caffe.coord_map import crop
def conv_relu(bottom, nout, ks=3, stride=1, pad=1):
conv = L.Convolution(bottom, kernel_size=ks, stride=stride,
num_output=nout, pad=pad,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
return conv, L.ReLU(conv, in_place=True)
def max_pool(bottom, ks=2, stride=2):
return L.Pooling(bottom, pool=P.Pooling.MAX, kernel_size=ks, stride=stride)
def fcn(split):
n = caffe.NetSpec()
pydata_params = dict(split=split, mean=(104.00699, 116.66877, 122.67892),
seed=1337)
if split == 'train':
pydata_params['sbdd_dir'] = './data/sbdd/dataset'
pylayer = 'SBDDSegDataLayer'
else:
pydata_params['voc_dir'] = '../data/pascal/VOC2011'
pylayer = 'VOCSegDataLayer'
n.data, n.label = L.Python(module='voc_layers', layer=pylayer,
ntop=2, param_str=str(pydata_params))
# the base net
n.conv1_1, n.relu1_1 = conv_relu(n.data, 64, pad=100)
n.conv1_2, n.relu1_2 = conv_relu(n.relu1_1, 64)
n.pool1 = max_pool(n.relu1_2)
n.conv2_1, n.relu2_1 = conv_relu(n.pool1, 128)
n.conv2_2, n.relu2_2 = conv_relu(n.relu2_1, 128)
n.pool2 = max_pool(n.relu2_2)
n.conv3_1, n.relu3_1 = conv_relu(n.pool2, 256)
n.conv3_2, n.relu3_2 = conv_relu(n.relu3_1, 256)
n.conv3_3, n.relu3_3 = conv_relu(n.relu3_2, 256)
n.pool3 = max_pool(n.relu3_3)
n.conv4_1, n.relu4_1 = conv_relu(n.pool3, 512)
n.conv4_2, n.relu4_2 = conv_relu(n.relu4_1, 512)
n.conv4_3, n.relu4_3 = conv_relu(n.relu4_2, 512)
n.pool4 = max_pool(n.relu4_3)
n.conv5_1, n.relu5_1 = conv_relu(n.pool4, 512)
n.conv5_2, n.relu5_2 = conv_relu(n.relu5_1, 512)
n.conv5_3, n.relu5_3 = conv_relu(n.relu5_2, 512)
n.pool5 = max_pool(n.relu5_3)
# fully conv
n.fc6, n.relu6 = conv_relu(n.pool5, 4096, ks=7, pad=0)
n.drop6 = L.Dropout(n.relu6, dropout_ratio=0.5, in_place=True)
n.fc7, n.relu7 = conv_relu(n.drop6, 4096, ks=1, pad=0)
n.drop7 = L.Dropout(n.relu7, dropout_ratio=0.5, in_place=True)
n.score_fr = L.Convolution(n.drop7, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.upscore2 = L.Deconvolution(n.score_fr,
convolution_param=dict(num_output=21, kernel_size=4, stride=2,
bias_term=False),
param=[dict(lr_mult=0)])
n.score_pool4 = L.Convolution(n.pool4, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.score_pool4c = crop(n.score_pool4, n.upscore2)
n.fuse_pool4 = L.Eltwise(n.upscore2, n.score_pool4c,
operation=P.Eltwise.SUM)
n.upscore16 = L.Deconvolution(n.fuse_pool4,
convolution_param=dict(num_output=21, kernel_size=32, stride=16,
bias_term=False),
param=[dict(lr_mult=0)])
n.score = crop(n.upscore16, n.data)
n.loss = L.SoftmaxWithLoss(n.score, n.label,
loss_param=dict(normalize=False, ignore_label=255))
return n.to_proto()
def make_net():
with open('train.prototxt', 'w') as f:
f.write(str(fcn('train')))
with open('val.prototxt', 'w') as f:
f.write(str(fcn('seg11valid')))
if __name__ == '__main__':
make_net()
examples/Seg-FCN/voc-fcn16s/solve.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Train a FCN-16s(PASCAL VOC) network """
import dragon.vm.caffe as caffe
import surgery
weights = '../voc-fcn32s/snapshot/train_iter_100000.caffemodel'
if __name__ == '__main__':
# init
caffe.set_mode_gpu()
caffe.set_device(0)
solver = caffe.SGDSolver('solver.prototxt')
solver.net.copy_from(weights)
# surgeries
interp_layers = [k for k in solver.net.params.keys() if 'up' in k]
surgery.interp(solver.net, interp_layers)
for _ in range(25):
solver.step(4000)
\ No newline at end of file
examples/Seg-FCN/voc-fcn16s/solver.prototxt 0 → 100644
train_net: "train.prototxt"
test_net: "val.prototxt"
test_iter: 1111
# make test net, but don't invoke it from the solver itself
test_interval: 999999999
display: 20
average_loss: 20
lr_policy: "fixed"
# lr for unnormalized softmax
base_lr: 1e-12
# high momentum
momentum: 0.99
# no gradient accumulation
iter_size: 1
max_iter: 100000
weight_decay: 0.0005
snapshot: 4000
snapshot_prefix: "snapshot/train"
test_initialization: false
examples/Seg-FCN/voc-fcn16s/test.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Test a FCN-16s(PASCAL VOC) network """
import dragon.vm.caffe as caffe
import score
import numpy as np
weights = 'snapshot/train_iter_44000.caffemodel'
if __name__ == '__main__':
# init
caffe.set_mode_gpu()
caffe.set_device(0)
solver = caffe.SGDSolver('solver.prototxt')
solver.net.copy_from(weights)
# scoring
val = np.loadtxt('../data/seg11valid.txt', dtype=str)
score.seg_tests(solver, 'seg', val)
examples/Seg-FCN/voc-fcn16s/train.prototxt 0 → 100644
layer {
name: "data"
type: "Python"
top: "data"
top: "label"
python_param {
module: "voc_layers"
layer: "SBDDSegDataLayer"
param_str: "{\'sbdd_dir\': \'./data/sbdd/dataset\', \'seed\': 1337, \'split\': \'train\', \'mean\': (104.00699, 116.66877, 122.67892)}"
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore2"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore2"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "score_pool4"
type: "Convolution"
bottom: "pool4"
top: "score_pool4"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool4c"
type: "Crop"
bottom: "score_pool4"
bottom: "upscore2"
top: "score_pool4c"
crop_param {
axis: 2
offset: 5
}
}
layer {
name: "fuse_pool4"
type: "Eltwise"
bottom: "upscore2"
bottom: "score_pool4c"
top: "fuse_pool4"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore16"
type: "Deconvolution"
bottom: "fuse_pool4"
top: "upscore16"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 32
stride: 16
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore16"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 27
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "score"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalize: false
}
}
examples/Seg-FCN/voc-fcn16s/val.prototxt 0 → 100644
layer {
name: "data"
type: "Python"
top: "data"
top: "label"
python_param {
module: "voc_layers"
layer: "VOCSegDataLayer"
param_str: "{\'voc_dir\': \'../data/pascal/VOC2011\', \'seed\': 1337, \'split\': \'seg11valid\', \'mean\': (104.00699, 116.66877, 122.67892)}"
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore2"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore2"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "score_pool4"
type: "Convolution"
bottom: "pool4"
top: "score_pool4"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool4c"
type: "Crop"
bottom: "score_pool4"
bottom: "upscore2"
top: "score_pool4c"
crop_param {
axis: 2
offset: 5
}
}
layer {
name: "fuse_pool4"
type: "Eltwise"
bottom: "upscore2"
bottom: "score_pool4c"
top: "fuse_pool4"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore16"
type: "Deconvolution"
bottom: "fuse_pool4"
top: "upscore16"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 32
stride: 16
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore16"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 27
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "score"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalize: false
}
}
examples/Seg-FCN/voc-fcn32s/caffemodel-url 0 → 100644
http://dl.caffe.berkeleyvision.org/fcn32s-heavy-pascal.caffemodel
\ No newline at end of file
examples/Seg-FCN/voc-fcn32s/net.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Source Code by Evan Shelhamer
# Re-Written by Ting Pan
# --------------------------------------------------------
import dragon.vm.caffe as caffe
from dragon.vm.caffe import layers as L, params as P
from dragon.vm.caffe.coord_map import crop
def conv_relu(bottom, nout, ks=3, stride=1, pad=1):
conv = L.Convolution(bottom, kernel_size=ks, stride=stride,
num_output=nout, pad=pad,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
return conv, L.ReLU(conv, in_place=True)
def max_pool(bottom, ks=2, stride=2):
return L.Pooling(bottom, pool=P.Pooling.MAX, kernel_size=ks, stride=stride)
def fcn(split):
n = caffe.NetSpec()
pydata_params = dict(split=split, mean=(104.00699, 116.66877, 122.67892),
seed=1337)
if split == 'train':
pydata_params['sbdd_dir'] = '../data/sbdd/dataset'
pylayer = 'SBDDSegDataLayer'
else:
pydata_params['voc_dir'] = '../data/pascal/VOC2011'
pylayer = 'VOCSegDataLayer'
n.data, n.label = L.Python(module='voc_layers', layer=pylayer,
ntop=2, param_str=str(pydata_params))
# the base net
n.conv1_1, n.relu1_1 = conv_relu(n.data, 64, pad=100)
n.conv1_2, n.relu1_2 = conv_relu(n.relu1_1, 64)
n.pool1 = max_pool(n.relu1_2)
n.conv2_1, n.relu2_1 = conv_relu(n.pool1, 128)
n.conv2_2, n.relu2_2 = conv_relu(n.relu2_1, 128)
n.pool2 = max_pool(n.relu2_2)
n.conv3_1, n.relu3_1 = conv_relu(n.pool2, 256)
n.conv3_2, n.relu3_2 = conv_relu(n.relu3_1, 256)
n.conv3_3, n.relu3_3 = conv_relu(n.relu3_2, 256)
n.pool3 = max_pool(n.relu3_3)
n.conv4_1, n.relu4_1 = conv_relu(n.pool3, 512)
n.conv4_2, n.relu4_2 = conv_relu(n.relu4_1, 512)
n.conv4_3, n.relu4_3 = conv_relu(n.relu4_2, 512)
n.pool4 = max_pool(n.relu4_3)
n.conv5_1, n.relu5_1 = conv_relu(n.pool4, 512)
n.conv5_2, n.relu5_2 = conv_relu(n.relu5_1, 512)
n.conv5_3, n.relu5_3 = conv_relu(n.relu5_2, 512)
n.pool5 = max_pool(n.relu5_3)
# fully conv
n.fc6, n.relu6 = conv_relu(n.pool5, 4096, ks=7, pad=0)
n.drop6 = L.Dropout(n.relu6, dropout_ratio=0.5, in_place=True)
n.fc7, n.relu7 = conv_relu(n.drop6, 4096, ks=1, pad=0)
n.drop7 = L.Dropout(n.relu7, dropout_ratio=0.5, in_place=True)
n.score_fr = L.Convolution(n.drop7, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.upscore = L.Deconvolution(n.score_fr,
convolution_param=dict(num_output=21, kernel_size=64, stride=32,
bias_term=False),
param=[dict(lr_mult=0)])
n.score = crop(n.upscore, n.data)
n.loss = L.SoftmaxWithLoss(n.score, n.label,
loss_param=dict(normalize=False, ignore_label=255))
return n.to_proto()
def make_net():
with open('train.prototxt', 'w') as f:
f.write(str(fcn('train')))
with open('val.prototxt', 'w') as f:
f.write(str(fcn('seg11valid')))
if __name__ == '__main__':
make_net()
examples/Seg-FCN/voc-fcn32s/solve.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Train a FCN-32s(PASCAL VOC) network """
import dragon.vm.caffe as caffe
import surgery
import numpy as np
weights = '../transplants/VGG16/VGG16.fcn.caffemodel'
if __name__ == '__main__':
# init
caffe.set_mode_gpu()
caffe.set_device(0)
solver = caffe.SGDSolver('solver.prototxt')
solver.net.copy_from(weights)
# surgeries
interp_layers = [k for k in solver.net.params.keys() if 'up' in k]
surgery.interp(solver.net, interp_layers)
for _ in range(25):
solver.step(4000)
\ No newline at end of file
examples/Seg-FCN/voc-fcn32s/solver.prototxt 0 → 100644
train_net: "train.prototxt"
test_net: "val.prototxt"
test_iter: 1111
# make test net, but don't invoke it from the solver itself
test_interval: 999999999
display: 20
average_loss: 20
lr_policy: "fixed"
# lr for unnormalized softmax
base_lr: 1e-10
# high momentum
momentum: 0.99
# no gradient accumulation
iter_size: 1
max_iter: 100000
weight_decay: 0.0005
snapshot: 4000
snapshot_prefix: "snapshot/train"
test_initialization: false
examples/Seg-FCN/voc-fcn32s/test.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Test a FCN-32s(PASCAL VOC) network """
import dragon.vm.caffe as caffe
import score
import numpy as np
weights = 'snapshot/train_iter_100000.caffemodel'
if __name__ == '__main__':
# init
caffe.set_mode_gpu()
caffe.set_device(0)
solver = caffe.SGDSolver('solver.prototxt')
solver.net.copy_from(weights)
# scoring
val = np.loadtxt('../data/seg11valid.txt', dtype=str)
score.seg_tests(solver, 'seg', val)
examples/Seg-FCN/voc-fcn32s/train.prototxt 0 → 100644
layer {
name: "data"
type: "Python"
top: "data"
top: "label"
python_param {
module: "voc_layers"
layer: "SBDDSegDataLayer"
param_str: "{\'sbdd_dir\': \'../data/sbdd/dataset\', \'seed\': 1337, \'split\': \'train\', \'mean\': (104.00699, 116.66877, 122.67892)}"
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 64
stride: 32
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 19
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "score"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalize: false
}
}
examples/Seg-FCN/voc-fcn32s/val.prototxt 0 → 100644
layer {
name: "data"
type: "Python"
top: "data"
top: "label"
python_param {
module: "voc_layers"
layer: "VOCSegDataLayer"
param_str: "{\'voc_dir\': \'../data/pascal/VOC2011\', \'seed\': 1337, \'split\': \'seg11valid\', \'mean\': (104.00699, 116.66877, 122.67892)}"
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 64
stride: 32
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 19
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "score"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalize: false
}
}
examples/Seg-FCN/voc-fcn8s-atonce/caffemodel-url 0 → 100644
http://dl.caffe.berkeleyvision.org/fcn8s-atonce-pascal.caffemodel
examples/Seg-FCN/voc-fcn8s-atonce/net.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Source Code by Evan Shelhamer
# Re-Written by Ting Pan
# --------------------------------------------------------
import dragon.vm.caffe as caffe
from dragon.vm.caffe import layers as L, params as P
from dragon.vm.caffe.coord_map import crop
def conv_relu(bottom, nout, ks=3, stride=1, pad=1):
conv = L.Convolution(bottom, kernel_size=ks, stride=stride,
num_output=nout, pad=pad,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
return conv, L.ReLU(conv, in_place=True)
def max_pool(bottom, ks=2, stride=2):
return L.Pooling(bottom, pool=P.Pooling.MAX, kernel_size=ks, stride=stride)
def fcn(split):
n = caffe.NetSpec()
pydata_params = dict(split=split, mean=(104.00699, 116.66877, 122.67892),
seed=1337)
if split == 'train':
pydata_params['sbdd_dir'] = '../data/sbdd/dataset'
pylayer = 'SBDDSegDataLayer'
else:
pydata_params['voc_dir'] = '../data/pascal/VOC2011'
pylayer = 'VOCSegDataLayer'
n.data, n.label = L.Python(module='voc_layers', layer=pylayer,
ntop=2, param_str=str(pydata_params))
# the base net
n.conv1_1, n.relu1_1 = conv_relu(n.data, 64, pad=100)
n.conv1_2, n.relu1_2 = conv_relu(n.relu1_1, 64)
n.pool1 = max_pool(n.relu1_2)
n.conv2_1, n.relu2_1 = conv_relu(n.pool1, 128)
n.conv2_2, n.relu2_2 = conv_relu(n.relu2_1, 128)
n.pool2 = max_pool(n.relu2_2)
n.conv3_1, n.relu3_1 = conv_relu(n.pool2, 256)
n.conv3_2, n.relu3_2 = conv_relu(n.relu3_1, 256)
n.conv3_3, n.relu3_3 = conv_relu(n.relu3_2, 256)
n.pool3 = max_pool(n.relu3_3)
n.conv4_1, n.relu4_1 = conv_relu(n.pool3, 512)
n.conv4_2, n.relu4_2 = conv_relu(n.relu4_1, 512)
n.conv4_3, n.relu4_3 = conv_relu(n.relu4_2, 512)
n.pool4 = max_pool(n.relu4_3)
n.conv5_1, n.relu5_1 = conv_relu(n.pool4, 512)
n.conv5_2, n.relu5_2 = conv_relu(n.relu5_1, 512)
n.conv5_3, n.relu5_3 = conv_relu(n.relu5_2, 512)
n.pool5 = max_pool(n.relu5_3)
# fully conv
n.fc6, n.relu6 = conv_relu(n.pool5, 4096, ks=7, pad=0)
n.drop6 = L.Dropout(n.relu6, dropout_ratio=0.5, in_place=True)
n.fc7, n.relu7 = conv_relu(n.drop6, 4096, ks=1, pad=0)
n.drop7 = L.Dropout(n.relu7, dropout_ratio=0.5, in_place=True)
n.score_fr = L.Convolution(n.drop7, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.upscore2 = L.Deconvolution(n.score_fr,
convolution_param=dict(num_output=21, kernel_size=4, stride=2,
bias_term=False),
param=[dict(lr_mult=0)])
# scale pool4 skip for compatibility
n.scale_pool4 = L.Scale(n.pool4, filler=dict(type='constant',
value=0.01), param=[dict(lr_mult=0)])
n.score_pool4 = L.Convolution(n.scale_pool4, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.score_pool4c = crop(n.score_pool4, n.upscore2)
n.fuse_pool4 = L.Eltwise(n.upscore2, n.score_pool4c,
operation=P.Eltwise.SUM)
n.upscore_pool4 = L.Deconvolution(n.fuse_pool4,
convolution_param=dict(num_output=21, kernel_size=4, stride=2,
bias_term=False),
param=[dict(lr_mult=0)])
# scale pool3 skip for compatibility
n.scale_pool3 = L.Scale(n.pool3, filler=dict(type='constant',
value=0.0001), param=[dict(lr_mult=0)])
n.score_pool3 = L.Convolution(n.scale_pool3, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.score_pool3c = crop(n.score_pool3, n.upscore_pool4)
n.fuse_pool3 = L.Eltwise(n.upscore_pool4, n.score_pool3c,
operation=P.Eltwise.SUM)
n.upscore8 = L.Deconvolution(n.fuse_pool3,
convolution_param=dict(num_output=21, kernel_size=16, stride=8,
bias_term=False),
param=[dict(lr_mult=0)])
n.score = crop(n.upscore8, n.data)
n.loss = L.SoftmaxWithLoss(n.score, n.label,
loss_param=dict(normalize=False, ignore_label=255))
return n.to_proto()
def make_net():
with open('train.prototxt', 'w') as f:
f.write(str(fcn('train')))
with open('val.prototxt', 'w') as f:
f.write(str(fcn('seg11valid')))
if __name__ == '__main__':
make_net()
examples/Seg-FCN/voc-fcn8s-atonce/solve.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Train a FCN-8s At Once(PASCAL VOC) network """
import dragon.vm.caffe as caffe
import surgery
weights = '../transplants/VGG16/VGG16.fcn.caffemodel'
if __name__ == '__main__':
# init
caffe.set_mode_gpu()
caffe.set_device(0)
solver = caffe.SGDSolver('solver.prototxt')
solver.net.copy_from(weights)
# surgeries
interp_layers = [k for k in solver.net.params.keys() if 'up' in k]
surgery.interp(solver.net, interp_layers)
for _ in range(75):
solver.step(4000)
\ No newline at end of file
examples/Seg-FCN/voc-fcn8s-atonce/solver.prototxt 0 → 100644
train_net: "train.prototxt"
test_net: "val.prototxt"
test_iter: 736
# make test net, but don't invoke it from the solver itself
test_interval: 999999999
display: 20
average_loss: 20
lr_policy: "fixed"
# lr for unnormalized softmax
base_lr: 1e-10
# high momentum
momentum: 0.99
# no gradient accumulation
iter_size: 1
max_iter: 300000
weight_decay: 0.0005
snapshot: 4000
snapshot_prefix: "snapshot/train"
test_initialization: false
examples/Seg-FCN/voc-fcn8s-atonce/test.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Test a FCN-8s At Once(PASCAL VOC) network """
import dragon.vm.caffe as caffe
import score
import numpy as np
weights = 'snapshot/train_iter_300000.caffemodel'
if __name__ == '__main__':
# init
caffe.set_mode_gpu()
caffe.set_device(0)
solver = caffe.SGDSolver('solver.prototxt')
solver.net.copy_from(weights)
# scoring
val = np.loadtxt('../data/seg11valid.txt', dtype=str)
score.seg_tests(solver, 'seg', val)
examples/Seg-FCN/voc-fcn8s-atonce/train.prototxt 0 → 100644
layer {
name: "data"
type: "Python"
top: "data"
top: "label"
python_param {
module: "voc_layers"
layer: "SBDDSegDataLayer"
param_str: "{\'sbdd_dir\': \'../data/sbdd/dataset\', \'seed\': 1337, \'split\': \'train\', \'mean\': (104.00699, 116.66877, 122.67892)}"
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore2"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore2"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "scale_pool4"
type: "Scale"
bottom: "pool4"
top: "scale_pool4"
param {
lr_mult: 0.0
}
scale_param {
filler {
type: "constant"
value: 0.00999999977648
}
}
}
layer {
name: "score_pool4"
type: "Convolution"
bottom: "scale_pool4"
top: "score_pool4"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool4c"
type: "Crop"
bottom: "score_pool4"
bottom: "upscore2"
top: "score_pool4c"
crop_param {
axis: 2
offset: 5
}
}
layer {
name: "fuse_pool4"
type: "Eltwise"
bottom: "upscore2"
bottom: "score_pool4c"
top: "fuse_pool4"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore_pool4"
type: "Deconvolution"
bottom: "fuse_pool4"
top: "upscore_pool4"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "scale_pool3"
type: "Scale"
bottom: "pool3"
top: "scale_pool3"
param {
lr_mult: 0.0
}
scale_param {
filler {
type: "constant"
value: 9.99999974738e-05
}
}
}
layer {
name: "score_pool3"
type: "Convolution"
bottom: "scale_pool3"
top: "score_pool3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool3c"
type: "Crop"
bottom: "score_pool3"
bottom: "upscore_pool4"
top: "score_pool3c"
crop_param {
axis: 2
offset: 9
}
}
layer {
name: "fuse_pool3"
type: "Eltwise"
bottom: "upscore_pool4"
bottom: "score_pool3c"
top: "fuse_pool3"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore8"
type: "Deconvolution"
bottom: "fuse_pool3"
top: "upscore8"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 16
stride: 8
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore8"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 31
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "score"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalize: false
}
}
examples/Seg-FCN/voc-fcn8s-atonce/val.prototxt 0 → 100644
layer {
name: "data"
type: "Python"
top: "data"
top: "label"
python_param {
module: "voc_layers"
layer: "VOCSegDataLayer"
param_str: "{\'voc_dir\': \'../data/pascal/VOC2011\', \'seed\': 1337, \'split\': \'seg11valid\', \'mean\': (104.00699, 116.66877, 122.67892)}"
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore2"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore2"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "scale_pool4"
type: "Scale"
bottom: "pool4"
top: "scale_pool4"
param {
lr_mult: 0.0
}
scale_param {
filler {
type: "constant"
value: 0.00999999977648
}
}
}
layer {
name: "score_pool4"
type: "Convolution"
bottom: "scale_pool4"
top: "score_pool4"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool4c"
type: "Crop"
bottom: "score_pool4"
bottom: "upscore2"
top: "score_pool4c"
crop_param {
axis: 2
offset: 5
}
}
layer {
name: "fuse_pool4"
type: "Eltwise"
bottom: "upscore2"
bottom: "score_pool4c"
top: "fuse_pool4"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore_pool4"
type: "Deconvolution"
bottom: "fuse_pool4"
top: "upscore_pool4"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "scale_pool3"
type: "Scale"
bottom: "pool3"
top: "scale_pool3"
param {
lr_mult: 0.0
}
scale_param {
filler {
type: "constant"
value: 9.99999974738e-05
}
}
}
layer {
name: "score_pool3"
type: "Convolution"
bottom: "scale_pool3"
top: "score_pool3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool3c"
type: "Crop"
bottom: "score_pool3"
bottom: "upscore_pool4"
top: "score_pool3c"
crop_param {
axis: 2
offset: 9
}
}
layer {
name: "fuse_pool3"
type: "Eltwise"
bottom: "upscore_pool4"
bottom: "score_pool3c"
top: "fuse_pool3"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore8"
type: "Deconvolution"
bottom: "fuse_pool3"
top: "upscore8"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 16
stride: 8
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore8"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 31
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "score"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalize: false
}
}
examples/Seg-FCN/voc-fcn8s/caffemodel-url 0 → 100644
http://dl.caffe.berkeleyvision.org/fcn8s-heavy-pascal.caffemodel
\ No newline at end of file
examples/Seg-FCN/voc-fcn8s/deploy.prototxt 0 → 100644
name: 'FCN-8s'
input: 'data'
input_dim: 1
input_dim: 3
input_dim: 224
input_dim: 224
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore2"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore2"
param {
lr_mult: 0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "score_pool4"
type: "Convolution"
bottom: "pool4"
top: "score_pool4"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool4c"
type: "Crop"
bottom: "score_pool4"
bottom: "upscore2"
top: "score_pool4c"
crop_param {
axis: 2
offset: 5
}
}
layer {
name: "fuse_pool4"
type: "Eltwise"
bottom: "upscore2"
bottom: "score_pool4c"
top: "fuse_pool4"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore_pool4"
type: "Deconvolution"
bottom: "fuse_pool4"
top: "upscore_pool4"
param {
lr_mult: 0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "score_pool3"
type: "Convolution"
bottom: "pool3"
top: "score_pool3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool3c"
type: "Crop"
bottom: "score_pool3"
bottom: "upscore_pool4"
top: "score_pool3c"
crop_param {
axis: 2
offset: 9
}
}
layer {
name: "fuse_pool3"
type: "Eltwise"
bottom: "upscore_pool4"
bottom: "score_pool3c"
top: "fuse_pool3"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore8"
type: "Deconvolution"
bottom: "fuse_pool3"
top: "upscore8"
param {
lr_mult: 0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 16
stride: 8
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore8"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 31
}
}
examples/Seg-FCN/voc-fcn8s/net.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Source Code by Evan Shelhamer
# Re-Written by Ting Pan
# --------------------------------------------------------
import dragon.vm.caffe as caffe
from dragon.vm.caffe import layers as L, params as P
from dragon.vm.caffe.coord_map import crop
def conv_relu(bottom, nout, ks=3, stride=1, pad=1):
conv = L.Convolution(bottom, kernel_size=ks, stride=stride,
num_output=nout, pad=pad,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
return conv, L.ReLU(conv, in_place=True)
def max_pool(bottom, ks=2, stride=2):
return L.Pooling(bottom, pool=P.Pooling.MAX, kernel_size=ks, stride=stride)
def fcn(split):
n = caffe.NetSpec()
pydata_params = dict(split=split, mean=(104.00699, 116.66877, 122.67892),
seed=1337)
if split == 'train':
pydata_params['sbdd_dir'] = '../data/sbdd/dataset'
pylayer = 'SBDDSegDataLayer'
else:
pydata_params['voc_dir'] = '../data/pascal/VOC2011'
pylayer = 'VOCSegDataLayer'
n.data, n.label = L.Python(module='voc_layers', layer=pylayer,
ntop=2, param_str=str(pydata_params))
# the base net
n.conv1_1, n.relu1_1 = conv_relu(n.data, 64, pad=100)
n.conv1_2, n.relu1_2 = conv_relu(n.relu1_1, 64)
n.pool1 = max_pool(n.relu1_2)
n.conv2_1, n.relu2_1 = conv_relu(n.pool1, 128)
n.conv2_2, n.relu2_2 = conv_relu(n.relu2_1, 128)
n.pool2 = max_pool(n.relu2_2)
n.conv3_1, n.relu3_1 = conv_relu(n.pool2, 256)
n.conv3_2, n.relu3_2 = conv_relu(n.relu3_1, 256)
n.conv3_3, n.relu3_3 = conv_relu(n.relu3_2, 256)
n.pool3 = max_pool(n.relu3_3)
n.conv4_1, n.relu4_1 = conv_relu(n.pool3, 512)
n.conv4_2, n.relu4_2 = conv_relu(n.relu4_1, 512)
n.conv4_3, n.relu4_3 = conv_relu(n.relu4_2, 512)
n.pool4 = max_pool(n.relu4_3)
n.conv5_1, n.relu5_1 = conv_relu(n.pool4, 512)
n.conv5_2, n.relu5_2 = conv_relu(n.relu5_1, 512)
n.conv5_3, n.relu5_3 = conv_relu(n.relu5_2, 512)
n.pool5 = max_pool(n.relu5_3)
# fully conv
n.fc6, n.relu6 = conv_relu(n.pool5, 4096, ks=7, pad=0)
n.drop6 = L.Dropout(n.relu6, dropout_ratio=0.5, in_place=True)
n.fc7, n.relu7 = conv_relu(n.drop6, 4096, ks=1, pad=0)
n.drop7 = L.Dropout(n.relu7, dropout_ratio=0.5, in_place=True)
n.score_fr = L.Convolution(n.drop7, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.upscore2 = L.Deconvolution(n.score_fr,
convolution_param=dict(num_output=21, kernel_size=4, stride=2,
bias_term=False),
param=[dict(lr_mult=0)])
n.score_pool4 = L.Convolution(n.pool4, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.score_pool4c = crop(n.score_pool4, n.upscore2)
n.fuse_pool4 = L.Eltwise(n.upscore2, n.score_pool4c,
operation=P.Eltwise.SUM)
n.upscore_pool4 = L.Deconvolution(n.fuse_pool4,
convolution_param=dict(num_output=21, kernel_size=4, stride=2,
bias_term=False),
param=[dict(lr_mult=0)])
n.score_pool3 = L.Convolution(n.pool3, num_output=21, kernel_size=1, pad=0,
param=[dict(lr_mult=1, decay_mult=1), dict(lr_mult=2, decay_mult=0)])
n.score_pool3c = crop(n.score_pool3, n.upscore_pool4)
n.fuse_pool3 = L.Eltwise(n.upscore_pool4, n.score_pool3c,
operation=P.Eltwise.SUM)
n.upscore8 = L.Deconvolution(n.fuse_pool3,
convolution_param=dict(num_output=21, kernel_size=16, stride=8,
bias_term=False),
param=[dict(lr_mult=0)])
n.score = crop(n.upscore8, n.data)
n.loss = L.SoftmaxWithLoss(n.score, n.label,
loss_param=dict(normalize=False, ignore_label=255))
return n.to_proto()
def make_net():
with open('train.prototxt', 'w') as f:
f.write(str(fcn('train')))
with open('val.prototxt', 'w') as f:
f.write(str(fcn('seg11valid')))
if __name__ == '__main__':
make_net()
examples/Seg-FCN/voc-fcn8s/solve.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Train a FCN-8s(PASCAL VOC) network """
import dragon.vm.caffe as caffe
import surgery
weights = '../voc-fcn16s/snapshot/train_iter_100000.caffemodel'
if __name__ == '__main__':
# init
caffe.set_mode_gpu()
caffe.set_device(0)
solver = caffe.SGDSolver('solver.prototxt')
solver.net.copy_from(weights)
# surgeries
interp_layers = [k for k in solver.net.params.keys() if 'up' in k]
surgery.interp(solver.net, interp_layers)
for _ in range(25):
solver.step(4000)
\ No newline at end of file
examples/Seg-FCN/voc-fcn8s/solver.prototxt 0 → 100644
train_net: "train.prototxt"
test_net: "val.prototxt"
test_iter: 1111
# make test net, but don't invoke it from the solver itself
test_interval: 999999999
display: 20
average_loss: 20
lr_policy: "fixed"
# lr for unnormalized softmax
base_lr: 1e-14
# high momentum
momentum: 0.99
# no gradient accumulation
iter_size: 1
max_iter: 100000
weight_decay: 0.0005
snapshot: 4000
snapshot_prefix: "snapshot/train"
test_initialization: false
examples/Seg-FCN/voc-fcn8s/test.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Written by Ting Pan
# --------------------------------------------------------
""" Test a FCN-8s(PASCAL VOC) network """
import dragon.vm.caffe as caffe
import score
import numpy as np
weights = 'snapshot/train_iter_100000.caffemodel'
if __name__ == '__main__':
# init
caffe.set_mode_gpu()
caffe.set_device(0)
solver = caffe.SGDSolver('solver.prototxt')
solver.net.copy_from(weights)
# scoring
val = np.loadtxt('../data/seg11valid.txt', dtype=str)
score.seg_tests(solver, 'D:/seg', val)
examples/Seg-FCN/voc-fcn8s/train.prototxt 0 → 100644
layer {
name: "data"
type: "Python"
top: "data"
top: "label"
python_param {
module: "voc_layers"
layer: "SBDDSegDataLayer"
param_str: "{\'sbdd_dir\': \'../data/sbdd/dataset\', \'seed\': 1337, \'split\': \'train\', \'mean\': (104.00699, 116.66877, 122.67892)}"
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore2"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore2"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "score_pool4"
type: "Convolution"
bottom: "pool4"
top: "score_pool4"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool4c"
type: "Crop"
bottom: "score_pool4"
bottom: "upscore2"
top: "score_pool4c"
crop_param {
axis: 2
offset: 5
}
}
layer {
name: "fuse_pool4"
type: "Eltwise"
bottom: "upscore2"
bottom: "score_pool4c"
top: "fuse_pool4"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore_pool4"
type: "Deconvolution"
bottom: "fuse_pool4"
top: "upscore_pool4"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "score_pool3"
type: "Convolution"
bottom: "pool3"
top: "score_pool3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool3c"
type: "Crop"
bottom: "score_pool3"
bottom: "upscore_pool4"
top: "score_pool3c"
crop_param {
axis: 2
offset: 9
}
}
layer {
name: "fuse_pool3"
type: "Eltwise"
bottom: "upscore_pool4"
bottom: "score_pool3c"
top: "fuse_pool3"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore8"
type: "Deconvolution"
bottom: "fuse_pool3"
top: "upscore8"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 16
stride: 8
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore8"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 31
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "score"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalize: false
}
}
examples/Seg-FCN/voc-fcn8s/val.prototxt 0 → 100644
layer {
name: "data"
type: "Python"
top: "data"
top: "label"
python_param {
module: "voc_layers"
layer: "VOCSegDataLayer"
param_str: "{\'voc_dir\': \'../data/pascal/VOC2011\', \'seed\': 1337, \'split\': \'seg11valid\', \'mean\': (104.00699, 116.66877, 122.67892)}"
}
}
layer {
name: "conv1_1"
type: "Convolution"
bottom: "data"
top: "conv1_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 100
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_1"
type: "ReLU"
bottom: "conv1_1"
top: "conv1_1"
}
layer {
name: "conv1_2"
type: "Convolution"
bottom: "conv1_1"
top: "conv1_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 64
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu1_2"
type: "ReLU"
bottom: "conv1_2"
top: "conv1_2"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "conv1_2"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv2_1"
type: "Convolution"
bottom: "pool1"
top: "conv2_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_1"
type: "ReLU"
bottom: "conv2_1"
top: "conv2_1"
}
layer {
name: "conv2_2"
type: "Convolution"
bottom: "conv2_1"
top: "conv2_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 128
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu2_2"
type: "ReLU"
bottom: "conv2_2"
top: "conv2_2"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "conv2_2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv3_1"
type: "Convolution"
bottom: "pool2"
top: "conv3_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_1"
type: "ReLU"
bottom: "conv3_1"
top: "conv3_1"
}
layer {
name: "conv3_2"
type: "Convolution"
bottom: "conv3_1"
top: "conv3_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_2"
type: "ReLU"
bottom: "conv3_2"
top: "conv3_2"
}
layer {
name: "conv3_3"
type: "Convolution"
bottom: "conv3_2"
top: "conv3_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu3_3"
type: "ReLU"
bottom: "conv3_3"
top: "conv3_3"
}
layer {
name: "pool3"
type: "Pooling"
bottom: "conv3_3"
top: "pool3"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv4_1"
type: "Convolution"
bottom: "pool3"
top: "conv4_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_1"
type: "ReLU"
bottom: "conv4_1"
top: "conv4_1"
}
layer {
name: "conv4_2"
type: "Convolution"
bottom: "conv4_1"
top: "conv4_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_2"
type: "ReLU"
bottom: "conv4_2"
top: "conv4_2"
}
layer {
name: "conv4_3"
type: "Convolution"
bottom: "conv4_2"
top: "conv4_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu4_3"
type: "ReLU"
bottom: "conv4_3"
top: "conv4_3"
}
layer {
name: "pool4"
type: "Pooling"
bottom: "conv4_3"
top: "pool4"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv5_1"
type: "Convolution"
bottom: "pool4"
top: "conv5_1"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_1"
type: "ReLU"
bottom: "conv5_1"
top: "conv5_1"
}
layer {
name: "conv5_2"
type: "Convolution"
bottom: "conv5_1"
top: "conv5_2"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_2"
type: "ReLU"
bottom: "conv5_2"
top: "conv5_2"
}
layer {
name: "conv5_3"
type: "Convolution"
bottom: "conv5_2"
top: "conv5_3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 512
pad: 1
kernel_size: 3
stride: 1
}
}
layer {
name: "relu5_3"
type: "ReLU"
bottom: "conv5_3"
top: "conv5_3"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5_3"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "fc6"
type: "Convolution"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 7
stride: 1
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "Convolution"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 4096
pad: 0
kernel_size: 1
stride: 1
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "score_fr"
type: "Convolution"
bottom: "fc7"
top: "score_fr"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "upscore2"
type: "Deconvolution"
bottom: "score_fr"
top: "upscore2"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "score_pool4"
type: "Convolution"
bottom: "pool4"
top: "score_pool4"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool4c"
type: "Crop"
bottom: "score_pool4"
bottom: "upscore2"
top: "score_pool4c"
crop_param {
axis: 2
offset: 5
}
}
layer {
name: "fuse_pool4"
type: "Eltwise"
bottom: "upscore2"
bottom: "score_pool4c"
top: "fuse_pool4"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore_pool4"
type: "Deconvolution"
bottom: "fuse_pool4"
top: "upscore_pool4"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 4
stride: 2
}
}
layer {
name: "score_pool3"
type: "Convolution"
bottom: "pool3"
top: "score_pool3"
param {
lr_mult: 1.0
decay_mult: 1.0
}
param {
lr_mult: 2.0
decay_mult: 0.0
}
convolution_param {
num_output: 21
pad: 0
kernel_size: 1
}
}
layer {
name: "score_pool3c"
type: "Crop"
bottom: "score_pool3"
bottom: "upscore_pool4"
top: "score_pool3c"
crop_param {
axis: 2
offset: 9
}
}
layer {
name: "fuse_pool3"
type: "Eltwise"
bottom: "upscore_pool4"
bottom: "score_pool3c"
top: "fuse_pool3"
eltwise_param {
operation: SUM
}
}
layer {
name: "upscore8"
type: "Deconvolution"
bottom: "fuse_pool3"
top: "upscore8"
param {
lr_mult: 0.0
}
convolution_param {
num_output: 21
bias_term: false
kernel_size: 16
stride: 8
}
}
layer {
name: "score"
type: "Crop"
bottom: "upscore8"
bottom: "data"
top: "score"
crop_param {
axis: 2
offset: 31
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "score"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalize: false
}
}
examples/Seg-FCN/voc_layers.py 0 → 100644
# --------------------------------------------------------
# Seg-FCN for Dragon
# Copyright (c) 2017 SeetaTech
# Source Code by Evan Shelhamer
# Re-Written by Ting Pan
# --------------------------------------------------------
import dragon.vm.caffe as caffe
import dragon.core.workspace as ws
import numpy as np
from PIL import Image
import random
class VOCSegDataLayer(caffe.Layer):
"""
Load (input image, label image) pairs from PASCAL VOC
one-at-a-time while reshaping the net to preserve dimensions.
Use this to feed data to a fully convolutional network.
"""
def setup(self, bottom, top):
"""
Setup data layer according to parameters:
- voc_dir: path to PASCAL VOC year dir
- split: train / val / test
- mean: tuple of mean values to subtract
- randomize: load in random order (default: True)
- seed: seed for randomization (default: None / current time)
for PASCAL VOC semantic segmentation.
example
params = dict(voc_dir="/path/to/PASCAL/VOC2011",
mean=(104.00698793, 116.66876762, 122.67891434),
split="val")
"""
# config
params = eval(self.param_str)
self.voc_dir = params['voc_dir']
self.split = params['split']
self.mean = np.array(params['mean'])
self.random = params.get('randomize', True)
self.seed = params.get('seed', None)
# two tops: data and label
if len(top) != 2:
raise Exception("Need to define two tops: data and label.")
# data layers have no bottoms
if len(bottom) != 0:
raise Exception("Do not define a bottom.")
# load indices for images and labels
split_f = '../data/{}.txt'.format(self.split)
self.indices = open(split_f, 'r').read().splitlines()
self.idx = 0
# make eval deterministic
if 'train' not in self.split:
self.random = False
# randomization: seed and pick
if self.random:
random.seed(self.seed)
self.idx = random.randint(0, len(self.indices)-1)
def reshape(self, bottom, top):
# load image + label image pair
self.data = self.load_image(self.indices[self.idx])
self.label = self.load_label(self.indices[self.idx])
# reshape tops to fit (leading 1 is for batch dimension)
self.data = self.data.reshape(1, *self.data.shape)
self.label = self.label.reshape(1, *self.label.shape)
def forward(self, bottom, top):
# assign output
ws.FeedTensor(top[0], self.data)
ws.FeedTensor(top[1], self.label)
# pick next input
if self.random:
self.idx = random.randint(0, len(self.indices)-1)
else:
self.idx += 1
if self.idx == len(self.indices):
self.idx = 0
def backward(self, top, propagate_down, bottom):
pass
def load_image(self, idx):
"""
Load input image and preprocess for Caffe:
- cast to float
- switch channels RGB -> BGR
- subtract mean
- transpose to channel x height x width order
"""
im = Image.open('{}/JPEGImages/{}.jpg'.format(self.voc_dir, idx))
in_ = np.array(im, dtype=np.float32)
in_ = in_[:,:,::-1]
in_ -= self.mean
in_ = in_.transpose((2,0,1))
return in_
def load_label(self, idx):
"""
Load label image as 1 x height x width integer array of label indices.
The leading singleton dimension is required by the loss.
"""
im = Image.open('{}/SegmentationClass/{}.png'.format(self.voc_dir, idx))
label = np.array(im, dtype=np.float32)
label = label[np.newaxis, ...]
return label
class SBDDSegDataLayer(caffe.Layer):
"""
Load (input image, label image) pairs from the SBDD extended labeling
of PASCAL VOC for semantic segmentation
one-at-a-time while reshaping the net to preserve dimensions.
Use this to feed data to a fully convolutional network.
"""
def setup(self, bottom, top):
"""
Setup data layer according to parameters:
- sbdd_dir: path to SBDD `dataset` dir
- split: train / seg11valid
- mean: tuple of mean values to subtract
- randomize: load in random order (default: True)
- seed: seed for randomization (default: None / current time)
for SBDD semantic segmentation.
N.B.segv11alid is the set of segval11 that does not intersect with SBDD.
Find it here: https://gist.github.com/shelhamer/edb330760338892d511e.
example
params = dict(sbdd_dir="/path/to/SBDD/dataset",
mean=(104.00698793, 116.66876762, 122.67891434),
split="valid")
"""
# config
params = eval(self.param_str)
self.sbdd_dir = params['sbdd_dir']
self.split = params['split']
self.mean = np.array(params['mean'])
self.random = params.get('randomize', True)
self.seed = params.get('seed', None)
# two tops: data and label
if len(top) != 2:
raise Exception("Need to define two tops: data and label.")
# data layers have no bottoms
if len(bottom) != 0:
raise Exception("Do not define a bottom.")
# load indices for images and labels
split_f = '{}/{}.txt'.format(self.sbdd_dir,
self.split)
self.indices = open(split_f, 'r').read().splitlines()
self.idx = 0
# make eval deterministic
if 'train' not in self.split:
self.random = False
# randomization: seed and pick
if self.random:
random.seed(self.seed)
self.idx = random.randint(0, len(self.indices)-1)
def reshape(self, bottom, top):
# load image + label image pair
self.data = self.load_image(self.indices[self.idx])
self.label = self.load_label(self.indices[self.idx])
# reshape tops to fit (leading 1 is for batch dimension)
self.data = self.data.reshape(1, *self.data.shape)
self.label = self.label.reshape(1, *self.label.shape)
def forward(self, bottom, top):
# assign output
ws.FeedTensor(top[0], self.data)
ws.FeedTensor(top[1], self.label)
# pick next input
if self.random:
self.idx = random.randint(0, len(self.indices)-1)
else:
self.idx += 1
if self.idx == len(self.indices):
self.idx = 0
def backward(self, top, propagate_down, bottom):
pass
def load_image(self, idx):
"""
Load input image and preprocess for Caffe:
- cast to float
- switch channels RGB -> BGR
- subtract mean
- transpose to channel x height x width order
"""
im = Image.open('{}/img/{}.jpg'.format(self.sbdd_dir, idx))
in_ = np.array(im, dtype=np.float32)
in_ = in_[:,:,::-1]
in_ -= self.mean
in_ = in_.transpose((2,0,1))
return in_
def load_label(self, idx):
"""
Load label image as 1 x height x width integer array of label indices.
The leading singleton dimension is required by the loss.
"""
import scipy.io
mat = scipy.io.loadmat('{}/cls/{}.mat'.format(self.sbdd_dir, idx))
label = mat['GTcls'][0]['Segmentation'][0].astype(np.float32)
label = label[np.newaxis, ...]
return label
Markdown is supported
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!