platform/upstream/caffe.git
10 years agodefine fully-convolutional imagenet model
Evan Shelhamer [Wed, 11 Jun 2014 16:57:07 +0000 (09:57 -0700)]
define fully-convolutional imagenet model

10 years agosave from python for net surgery
Evan Shelhamer [Mon, 26 May 2014 06:49:51 +0000 (23:49 -0700)]
save from python for net surgery

0. Scheme desired parameters.
1. Do surgery on the net through `net.params['name'][idx].data[...] = `.
2. Save post-operation net params by `net.save('fname')`.

Handwoven deep nets, anyone?

10 years agoMerge pull request #469 from weinman/grayscale-io-convert
Evan Shelhamer [Wed, 11 Jun 2014 17:17:23 +0000 (10:17 -0700)]
Merge pull request #469 from weinman/grayscale-io-convert

Add grayscale input processing for intensity images in tools and
pycaffe.

10 years agopycaffe: leave grayscale images gray according to arg
Evan Shelhamer [Tue, 10 Jun 2014 22:01:41 +0000 (15:01 -0700)]
pycaffe: leave grayscale images gray according to arg

10 years agodrop learning rates and decays from deploy model
Evan Shelhamer [Wed, 11 Jun 2014 16:55:54 +0000 (09:55 -0700)]
drop learning rates and decays from deploy model

10 years agogroom install docs
Evan Shelhamer [Tue, 10 Jun 2014 21:46:53 +0000 (14:46 -0700)]
groom install docs

- make OS X boost compilation more clear
- make punctuation more sincere

10 years agofix clang compilation problem w/ DummyDataLayer
Jeff Donahue [Tue, 10 Jun 2014 18:35:08 +0000 (11:35 -0700)]
fix clang compilation problem w/ DummyDataLayer

10 years agolint dummy data layer
Jeff Donahue [Mon, 9 Jun 2014 18:52:06 +0000 (11:52 -0700)]
lint dummy data layer

10 years agoMerge pull request #480 from jeffdonahue/dummy-data-layer
Jeff Donahue [Mon, 9 Jun 2014 17:58:26 +0000 (10:58 -0700)]
Merge pull request #480 from jeffdonahue/dummy-data-layer

Dummy data layer

10 years agoadd DummyDataLayer tests
Jeff Donahue [Sun, 8 Jun 2014 23:04:41 +0000 (16:04 -0700)]
add DummyDataLayer tests

10 years agoadd DummyDataLayer
Jeff Donahue [Sun, 8 Jun 2014 21:58:53 +0000 (14:58 -0700)]
add DummyDataLayer

10 years agofix ArgMaxLayer bug in num bottom blobs decl. pointed out by @sguada
Jeff Donahue [Mon, 9 Jun 2014 05:44:32 +0000 (22:44 -0700)]
fix ArgMaxLayer bug in num bottom blobs decl. pointed out by @sguada

10 years agoMerge pull request #479 from jeffdonahue/declare-layer-names-and-numblobs
Evan Shelhamer [Mon, 9 Jun 2014 04:11:24 +0000 (21:11 -0700)]
Merge pull request #479 from jeffdonahue/declare-layer-names-and-numblobs

Layers declare their types and number of bottom/top blobs

10 years agomove MemoryDataLayer decl. from vision_layers.hpp to data_layers.hpp
Jeff Donahue [Sun, 8 Jun 2014 19:45:17 +0000 (12:45 -0700)]
move MemoryDataLayer decl. from vision_layers.hpp to data_layers.hpp

10 years agolayers declare their names and number of input/output blobs, and don't
Jeff Donahue [Fri, 6 Jun 2014 22:40:26 +0000 (15:40 -0700)]
layers declare their names and number of input/output blobs, and don't
check top/bottom blob counts explicitly in SetUp; instead call base
Layer::SetUp.

10 years agofix Makefile build dir link upgrade bug reported by @jamt9000
Jeff Donahue [Mon, 9 Jun 2014 01:08:24 +0000 (18:08 -0700)]
fix Makefile build dir link upgrade bug reported by @jamt9000

10 years agoChanged variable name: iscolor to is_color.
Jerod Weinman [Sun, 8 Jun 2014 20:46:23 +0000 (15:46 -0500)]
Changed variable name: iscolor to is_color.

10 years agoMerge pull request #473 from shelhamer/pad-max-pooling
Jeff Donahue [Sun, 8 Jun 2014 19:03:22 +0000 (12:03 -0700)]
Merge pull request #473 from shelhamer/pad-max-pooling

Padding for Max Pooling

10 years agocheck the last pooling in padding and add padded max pooling test
Jeff Donahue [Fri, 6 Jun 2014 20:59:19 +0000 (13:59 -0700)]
check the last pooling in padding and add padded max pooling test

[Adapted from a commit by @jeffdonahue by @shelhamer.]

10 years agopadding for max pooling
Evan Shelhamer [Thu, 5 Jun 2014 17:53:42 +0000 (10:53 -0700)]
padding for max pooling

Max pooling pads by -inf if the padding parameter is set.

Padding for pooling, like padding for convolution, can preserve the
dimensions of the bottom at the top. By setting the padding to
floor(kernel_size / 2) the top output is the "same" instead of the
"valid" part of the bottom input.

10 years agoMerge pull request #475 from jeffdonahue/tanh-fixes
Evan Shelhamer [Sun, 8 Jun 2014 05:00:05 +0000 (22:00 -0700)]
Merge pull request #475 from jeffdonahue/tanh-fixes

Make TanH cleaner, more efficient, and possible to use in-place

10 years agoMake TanH cleaner, more efficient, and possible to use in-place
Jeff Donahue [Fri, 6 Jun 2014 00:20:30 +0000 (17:20 -0700)]
Make TanH cleaner, more efficient, and possible to use in-place

10 years agoMerge pull request #471 from jeffdonahue/debug-release-build-dirs
Jeff Donahue [Fri, 6 Jun 2014 01:06:01 +0000 (18:06 -0700)]
Merge pull request #471 from jeffdonahue/debug-release-build-dirs

Compile debug/release into separate directories

10 years agoMerge pull request #466 from robcurrie/dev
Evan Shelhamer [Tue, 3 Jun 2014 13:21:22 +0000 (09:21 -0400)]
Merge pull request #466 from robcurrie/dev

Update docs on building boost on OSX for the python wrappers

10 years agoupdate .gitignore appropriately for separate debug/release build dirs
Jeff Donahue [Tue, 3 Jun 2014 06:57:37 +0000 (23:57 -0700)]
update .gitignore appropriately for separate debug/release build dirs

10 years agocompile debug/release into separate directories so you don't have to
Jeff Donahue [Tue, 3 Jun 2014 06:47:09 +0000 (23:47 -0700)]
compile debug/release into separate directories so you don't have to
rebuild the whole thing to switch back and forth

10 years agoAdded an iscolor flag to io.cpp method ReadImageToDatum to handle grayscale images...
Jerod Weinman [Mon, 2 Jun 2014 20:54:25 +0000 (15:54 -0500)]
Added an iscolor flag to io.cpp method ReadImageToDatum to handle grayscale images and a corresponding commandline flag [-g] to convert_imageset.cpp.

10 years agoUpdate docs on building boost on OSX for the python wrappers
Rob Currie [Thu, 29 May 2014 21:46:29 +0000 (14:46 -0700)]
Update docs on building boost on OSX for the python wrappers

10 years agofix OSX 10.9 homebrew CXX doc
Evan Shelhamer [Thu, 29 May 2014 00:18:22 +0000 (17:18 -0700)]
fix OSX 10.9 homebrew CXX doc

clang++ and not clang as accidentally committed in 2dcbcd9.

10 years agoMerge pull request #422 from sguada/threshold_layer
Sergio Guadarrama [Tue, 27 May 2014 19:06:57 +0000 (12:06 -0700)]
Merge pull request #422 from sguada/threshold_layer

Threshold layer to binarize features
Added GPU code and tested

10 years agoUn comment Test GPUs cases, fixed ThresholdLayer.cu
Sergio [Tue, 27 May 2014 18:54:53 +0000 (11:54 -0700)]
Un comment Test GPUs cases, fixed ThresholdLayer.cu

10 years agoComment Test GPUs cases
Sergio [Tue, 27 May 2014 18:44:17 +0000 (11:44 -0700)]
Comment Test GPUs cases

10 years agoMake lint happy
Sergio [Tue, 27 May 2014 18:06:56 +0000 (11:06 -0700)]
Make lint happy

10 years agoFixed call to ThresholdForward in ThresholdLayer.cu
Sergio [Tue, 27 May 2014 18:02:20 +0000 (11:02 -0700)]
Fixed call to ThresholdForward in ThresholdLayer.cu

10 years agoFixed type in ThresholdLayer.cu
Sergio [Tue, 27 May 2014 18:01:19 +0000 (11:01 -0700)]
Fixed type in ThresholdLayer.cu

10 years agoFixed typo in Threshold Layer definition
Sergio [Tue, 27 May 2014 17:59:36 +0000 (10:59 -0700)]
Fixed typo in Threshold Layer definition

10 years agoAdded ForwardGPU to ThresholdLayer and to the tests
Sergio [Tue, 27 May 2014 17:57:26 +0000 (10:57 -0700)]
Added ForwardGPU to ThresholdLayer and to the tests

10 years agoAdded Threshold layer to neuron_layers.hpp
Sergio [Tue, 27 May 2014 17:54:21 +0000 (10:54 -0700)]
Added Threshold layer to neuron_layers.hpp

10 years agoCorrected conditions in test_threshold
Sergio Guadarrama [Fri, 16 May 2014 01:58:04 +0000 (18:58 -0700)]
Corrected conditions in test_threshold

10 years agoFix typo in test_threshold ThresholdParameter
Sergio Guadarrama [Fri, 16 May 2014 01:47:57 +0000 (18:47 -0700)]
Fix typo in test_threshold ThresholdParameter

10 years agoAdded the code for threshold_layer to the repo
Sergio Guadarrama [Fri, 16 May 2014 01:42:50 +0000 (18:42 -0700)]
Added the code for threshold_layer to the repo

10 years agoAdded NeuronLayer<Dtype>::SetUp(bottom, top) to ThresholdLayer
Sergio Guadarrama [Fri, 16 May 2014 00:14:00 +0000 (17:14 -0700)]
Added NeuronLayer<Dtype>::SetUp(bottom, top) to ThresholdLayer

10 years agoAdded threshold setting test
Sergio Guadarrama [Fri, 16 May 2014 00:01:50 +0000 (17:01 -0700)]
Added threshold setting test

10 years agoFixed name of blob_bottom_
Sergio Guadarrama [Thu, 15 May 2014 23:58:27 +0000 (16:58 -0700)]
Fixed name of blob_bottom_

10 years agoFixed name of threshold_ var
Sergio Guadarrama [Thu, 15 May 2014 23:57:50 +0000 (16:57 -0700)]
Fixed name of threshold_ var

10 years agoFixed ThresholdParam
Sergio [Tue, 27 May 2014 17:32:11 +0000 (10:32 -0700)]
Fixed ThresholdParam

Conflicts:

src/caffe/proto/caffe.proto

Conflicts:

src/caffe/proto/caffe.proto

Conflicts:
src/caffe/proto/caffe.proto

10 years agoTest for Threshold layer
Sergio [Thu, 15 May 2014 16:30:07 +0000 (09:30 -0700)]
Test for Threshold layer

10 years agoMerge pull request #459 from shelhamer/python-net-preprocessing-members
Evan Shelhamer [Tue, 27 May 2014 04:58:34 +0000 (21:58 -0700)]
Merge pull request #459 from shelhamer/python-net-preprocessing-members

Make net preprocessing options belong to instantiated net and not class

10 years agoMerge pull request #459 from shelhamer/python-net-preprocessing-members
Evan Shelhamer [Tue, 27 May 2014 04:58:34 +0000 (21:58 -0700)]
Merge pull request #459 from shelhamer/python-net-preprocessing-members

Make net preprocessing options belong to instantiated net and not class

10 years agocaffe.Net preprocessing members belong to object, not class
Evan Shelhamer [Tue, 27 May 2014 04:50:39 +0000 (21:50 -0700)]
caffe.Net preprocessing members belong to object, not class

10 years agoMerge pull request #445 from jeffdonahue/convert_imageset_resize_option
Evan Shelhamer [Mon, 26 May 2014 17:50:09 +0000 (10:50 -0700)]
Merge pull request #445 from jeffdonahue/convert_imageset_resize_option

Optionally resize an image set to canonical dimensions when converting.
Do this by default for creating the imagenet train + val leveldbs.

10 years agoconvert imageset comment fixup
Evan Shelhamer [Mon, 26 May 2014 17:50:00 +0000 (10:50 -0700)]
convert imageset comment fixup

10 years agoMerge pull request #456 from longjon/spurious-ldflags
Evan Shelhamer [Mon, 26 May 2014 16:54:20 +0000 (09:54 -0700)]
Merge pull request #456 from longjon/spurious-ldflags

Don't pass LDFLAGS when only performing compilation (-c)

10 years agodon't pass LDFLAGS when only compiling
Jonathan L Long [Mon, 26 May 2014 08:34:32 +0000 (01:34 -0700)]
don't pass LDFLAGS when only compiling

10 years ago10.9 install doc formatting
Evan Shelhamer [Sun, 25 May 2014 23:19:14 +0000 (16:19 -0700)]
10.9 install doc formatting

10 years agoBack-merge recent fixes from master to dev
Evan Shelhamer [Sun, 25 May 2014 04:41:16 +0000 (21:41 -0700)]
Back-merge recent fixes from master to dev

fix OSX 10.9 compiler/stdlib override for latest homebrew
follow-up on #443 to invert k channels (instead of 3)
Correctly invert the swapping of colour channels
link presentation on dropbox (was self-hosted during a dropbox issue)
link to demo
fix draw_net python script
release v1 model defs + weights
point out @niuzhiheng's work on the Windows port
fix test_all path in docs
link canonical bvlc site
fix detection notebook link

10 years agofix OSX 10.9 compiler/stdlib override for latest homebrew
Evan Shelhamer [Sun, 25 May 2014 04:06:37 +0000 (21:06 -0700)]
fix OSX 10.9 compiler/stdlib override for latest homebrew

10 years agoMerge pull request #448 from jeffdonahue/sguada-fix_maxpooling
Sergio Guadarrama [Sun, 25 May 2014 02:39:58 +0000 (19:39 -0700)]
Merge pull request #448 from jeffdonahue/sguada-fix_maxpooling

Finish up max pooling with a mask from @sguada
Tested

10 years agomerge caffe_set definitions; define for int as well
Jeff Donahue [Sun, 25 May 2014 02:02:51 +0000 (19:02 -0700)]
merge caffe_set definitions; define for int as well

10 years agoadd tests for maxpooling layer forward, and for maxpooling with top mask
Jeff Donahue [Sun, 25 May 2014 01:40:03 +0000 (18:40 -0700)]
add tests for maxpooling layer forward, and for maxpooling with top mask

10 years agooptionally output the mask to a top blob instead of storing internally
Jeff Donahue [Sun, 25 May 2014 01:09:05 +0000 (18:09 -0700)]
optionally output the mask to a top blob instead of storing internally

10 years agomake a Blob<unsigned int> and use in dropout layer
Jeff Donahue [Sat, 24 May 2014 23:59:15 +0000 (16:59 -0700)]
make a Blob<unsigned int> and use in dropout layer

10 years agouse a Blob<int> instead of a SyncedMemory to store max_idx_
Jeff Donahue [Sat, 24 May 2014 23:27:38 +0000 (16:27 -0700)]
use a Blob<int> instead of a SyncedMemory to store max_idx_

10 years agobugfix: setting count to the top count in backward doesn't process all
Jeff Donahue [Sat, 24 May 2014 23:14:44 +0000 (16:14 -0700)]
bugfix: setting count to the top count in backward doesn't process all
of the bottom (assuming the bottom is larger, which happens for
nontrivial poolsize>1)

10 years agomask should be const in backward pass
Jeff Donahue [Sat, 24 May 2014 23:14:04 +0000 (16:14 -0700)]
mask should be const in backward pass

10 years agoremove commented out code
Jeff Donahue [Sat, 24 May 2014 23:10:23 +0000 (16:10 -0700)]
remove commented out code

10 years agolint and make compilable (using static_cast's found a couple bugs at
Jeff Donahue [Sat, 24 May 2014 22:29:35 +0000 (15:29 -0700)]
lint and make compilable (using static_cast's found a couple bugs at
compile time)

10 years agoAdapted to V1 proto definition, test don't pass
Sergio [Mon, 14 Apr 2014 04:03:06 +0000 (21:03 -0700)]
Adapted to V1 proto definition, test don't pass

10 years agoCommented Atomic Add, back to loop in GPU MaxPoolBackward
Sergio [Mon, 14 Apr 2014 03:29:26 +0000 (20:29 -0700)]
Commented Atomic Add, back to loop in GPU MaxPoolBackward

10 years agoAttempt to use AtomicAdd but it seems slower
Sergio [Mon, 14 Apr 2014 03:27:16 +0000 (20:27 -0700)]
Attempt to use AtomicAdd but it seems slower

10 years agoAdded test for maxpool layer followed by dropout
Sergio [Wed, 26 Feb 2014 01:29:43 +0000 (17:29 -0800)]
Added test for maxpool layer followed by dropout

10 years agoUse loops in GPU again to avoid over-writting of bottom_diff
Sergio [Mon, 14 Apr 2014 03:21:15 +0000 (20:21 -0700)]
Use loops in GPU again to avoid over-writting of bottom_diff

10 years agoFixed parameter order
Sergio [Mon, 14 Apr 2014 03:14:43 +0000 (20:14 -0700)]
Fixed parameter order

10 years agoCleaned prints from test_pooling_layer.cpp
Sergio [Mon, 14 Apr 2014 03:13:05 +0000 (20:13 -0700)]
Cleaned prints from test_pooling_layer.cpp

10 years agoSet bottom_diff to 0 and remove Async memcopy
Sergio [Mon, 14 Apr 2014 03:11:49 +0000 (20:11 -0700)]
Set bottom_diff to 0 and remove Async memcopy

10 years agoRemove top_data from backward Max Pooling
Sergio [Mon, 14 Apr 2014 03:10:10 +0000 (20:10 -0700)]
Remove top_data from backward Max Pooling

10 years agoUse mask_idx to compute backward Max Pooling
Sergio [Mon, 14 Apr 2014 03:06:38 +0000 (20:06 -0700)]
Use mask_idx to compute backward Max Pooling

10 years agoAdded test for Pooling layer GPU
Sergio [Mon, 14 Apr 2014 03:02:34 +0000 (20:02 -0700)]
Added test for  Pooling layer GPU

10 years agoAdded max_idx to Pooling layer GPU
Sergio [Mon, 14 Apr 2014 02:58:05 +0000 (19:58 -0700)]
Added max_idx to Pooling layer GPU

10 years agoDefault mask idx is -1
Sergio Guadarrama [Tue, 25 Feb 2014 01:36:17 +0000 (17:36 -0800)]
Default mask idx is -1

10 years agoAdded max_idx to Pooling layer CPU
Sergio [Mon, 14 Apr 2014 02:49:02 +0000 (19:49 -0700)]
Added max_idx to Pooling layer CPU

10 years agoadd convert_imageset option to resize images; use in
Jeff Donahue [Fri, 23 May 2014 21:57:02 +0000 (14:57 -0700)]
add convert_imageset option to resize images; use in
convert_imageset.cpp and document

10 years agofollow-up on #443 to invert k channels (instead of 3)
Evan Shelhamer [Fri, 23 May 2014 16:13:44 +0000 (09:13 -0700)]
follow-up on #443 to invert k channels (instead of 3)

10 years agoCorrectly invert the swapping of colour channels
James Thewlis [Fri, 23 May 2014 14:59:52 +0000 (15:59 +0100)]
Correctly invert the swapping of colour channels

In the 'deprocess' method, get back the image with the original channel order
by inverting the original transform, rather than reversing the tuple which is
incorrect.

10 years agofollow-up on #443 to invert k channels (instead of 3)
Evan Shelhamer [Fri, 23 May 2014 16:13:44 +0000 (09:13 -0700)]
follow-up on #443 to invert k channels (instead of 3)

10 years agoMerge pull request #443 from jamt9000/correct-deprocess
Evan Shelhamer [Fri, 23 May 2014 16:02:43 +0000 (09:02 -0700)]
Merge pull request #443 from jamt9000/correct-deprocess

Correctly invert the swapping of colour channels in Python API 'deprocess'

10 years agoCorrectly invert the swapping of colour channels
James Thewlis [Fri, 23 May 2014 14:59:52 +0000 (15:59 +0100)]
Correctly invert the swapping of colour channels

In the 'deprocess' method, get back the image with the original channel order
by inverting the original transform, rather than reversing the tuple which is
incorrect.

10 years agoMerge pull request #433 from shelhamer/eltwise
Evan Shelhamer [Fri, 23 May 2014 04:27:54 +0000 (21:27 -0700)]
Merge pull request #433 from shelhamer/eltwise

Elementwise layer takes sum or product; caffe_gpu_{add,sub}

10 years agocommment, lint
Evan Shelhamer [Fri, 23 May 2014 04:24:16 +0000 (21:24 -0700)]
commment, lint

10 years agoweight elementwise sum with per-blob coefficients
Evan Shelhamer [Fri, 23 May 2014 03:12:11 +0000 (20:12 -0700)]
weight elementwise sum with per-blob coefficients

10 years agolink presentation on dropbox (was self-hosted during a dropbox issue)
Evan Shelhamer [Fri, 23 May 2014 01:24:09 +0000 (18:24 -0700)]
link presentation on dropbox (was self-hosted during a dropbox issue)

10 years agolink to demo
Sergey Karayev [Fri, 23 May 2014 01:18:25 +0000 (18:18 -0700)]
link to demo

10 years agomake sum the default eltwise operation
Evan Shelhamer [Fri, 23 May 2014 00:41:24 +0000 (17:41 -0700)]
make sum the default eltwise operation

10 years agofix layer name in logging
Evan Shelhamer [Thu, 22 May 2014 09:13:04 +0000 (02:13 -0700)]
fix layer name in logging

10 years agoMerge pull request #435 from shelhamer/v1-models
Evan Shelhamer [Thu, 22 May 2014 08:04:36 +0000 (01:04 -0700)]
Merge pull request #435 from shelhamer/v1-models

Release v1 model defs + weights

10 years agofix draw_net python script
Evan Shelhamer [Thu, 22 May 2014 08:02:56 +0000 (01:02 -0700)]
fix draw_net python script

include caffe.draw for drawing functions.

10 years agorelease v1 model defs + weights
Evan Shelhamer [Thu, 22 May 2014 07:56:35 +0000 (00:56 -0700)]
release v1 model defs + weights

- Caffe reference ImageNet model
- AlexNet

Note that one can upgrade the weights locally by
`upgrade_net_proto_binary.bin` to avoid re-downloading.

10 years agoMerge pull request #434 from shelhamer/little-cat
Evan Shelhamer [Thu, 22 May 2014 07:50:04 +0000 (00:50 -0700)]
Merge pull request #434 from shelhamer/little-cat

Reduce example image size

10 years agoreduce example image size
Evan Shelhamer [Thu, 22 May 2014 07:30:12 +0000 (00:30 -0700)]
reduce example image size

10 years agopoint out @niuzhiheng's work on the Windows port
Evan Shelhamer [Thu, 22 May 2014 06:43:49 +0000 (23:43 -0700)]
point out @niuzhiheng's work on the Windows port