Detection#13

Open

NourElmadany wants to merge 1092 commits intoMMLab-CU:masterfrom

yjxiong:untrimmednet

This pull request is big! We're only showing the most recent 250 commits

Commits on Jul 6, 2015

minor fixes
yjxiong
committed

Commits on Jul 7, 2015

List protobuf-compiler dependency in the correct place (it is in the package managers for both 14.04 and 12.04)
semitrivial
committed
imported video data layer
lmwang
committed

Commits on Jul 9, 2015

Commits on Jul 10, 2015

Update README.md
yjxiong
committed
Update README.md
yjxiong
committed
Update README.md
yjxiong
committed
Merge pull request #2712 from semitrivial/master
jeffdonahue
committed
experimental multi-scale cropping support.
yjxiong
committed
Merge branch 'action_recog' of https://github.com/yjxiong/caffe into action_recog
lmwang
committed
Merge remote-tracking branch 'origin/data_parallel' into action_recog
lmwang
committed
multi-scale bug fix
yjxiong
committed
Merge remote-tracking branch 'upstream/master'
yjxiong
committed
Merge branch 'master' into data_parallel
yjxiong
committed
multi-scale bug fix
yjxiong
committed

Commits on Jul 11, 2015

Merge remote-tracking branch 'origin/data_parallel' into action_recog
lmwang
committed
multi-scale bug fix
yjxiong
committed

Commits on Jul 12, 2015

add corner crop and multiscale
lmwang
committed
Merge remote-tracking branch 'origin/data_parallel' into action_recog
lmwang
committed
Update README.md
yjxiong
committed
remove redundant cropping settings.
yjxiong
committed
resolve conlict of multi-scale and mean file
yjxiong
committed

Commits on Jul 14, 2015

Commits on Jul 28, 2015

Merge remote-tracking branch 'origin/data_parallel' into action_recog
wanglimin
committed

Commits on Jul 29, 2015

add is_flow in data_tranformer layer
lmwang
committed
modified gather_layer
lmwang
committed
solve the confict
lmwang
committed

Commits on Jul 30, 2015

Commits on Jul 31, 2015

Commits on Aug 2, 2015

add example of training very deep two-stream cnns
lmwang
committed

Commits on Aug 3, 2015

Commits on Aug 6, 2015

Update vgg_16_rgb_train_val_fast.prototxt
Limin Wang
committed
Update vgg_16_flow_train_val_fast.prototxt
Limin Wang
committed

Commits on Aug 7, 2015

Update readme.md
yjxiong
committed

Commits on Aug 11, 2015

Commits on Aug 14, 2015

solve testing issues
yjxiong
committed
Merge branch 'data_parallel' into action_recog
yjxiong
committed
Update README.md
yjxiong
committed

Commits on Aug 19, 2015

experimental batch normalization
lxx1991
committed

Commits on Aug 26, 2015

initial implementation finished.
yjxiong
committed
background comm implemented. Incorrect perforamance.
yjxiong
committed
cp
yjxiong
committed

Commits on Aug 31, 2015

results corrected. locking conflict solved.
yjxiong
committed

Commits on Sep 1, 2015

Update vgg_16_rgb_train_val_fast.prototxt
Limin Wang
committed
possible memleak. iter_size issue
yjxiong
committed
Update vgg_16_flow_train_val_fast.prototxt
Limin Wang
committed
Update vgg_16_rgb_train_val_fast.prototxt
Limin Wang
committed
memleak resolved. wrapped code in USE_MPI macro
yjxiong
committed
remove redundant code
yjxiong
committed
no need for boost 1.55 for now
yjxiong
committed

Commits on Sep 3, 2015

mpi_thread_serialized works properly with current ver.
yjxiong
committed

Commits on Sep 7, 2015

Merge branch 'bg_comm' into data_parallel
yjxiong
committed
Squashed commit of the following:
yjxiong
committed
Merge remote-tracking branch 'origin/action_recog' into action_recog
yjxiong
committed
Update README.md
yjxiong
committed
Update README.md
yjxiong
committed

Commits on Sep 8, 2015

Update README.md
yjxiong
committed

Commits on Sep 9, 2015

Update data_transformer.cpp
Limin Wang
committed
Update data_transformer.cpp
Limin Wang
committed
Update README.md
yjxiong
committed

Commits on Sep 11, 2015

Update README.md
yjxiong
committed

Commits on Sep 14, 2015

fix Gather layer bug when USE_MPI is disabled.
yjxiong
committed
Merge branch 'data_parallel' into action_recog
yjxiong
committed
Merge remote-tracking branch 'origin/action_recog' into action_recog
yjxiong
committed

Commits on Sep 17, 2015

add "exp10" lr_policy
yjxiong
committed
within-shard shuffle implemented.
yjxiong
committed

Commits on Sep 18, 2015

minor fixes
yjxiong
committed
minor fixes
yjxiong
committed
lr_policy typo
yjxiong
committed
disable openmpi cuda api: occasional data corruption.
yjxiong
committed

Commits on Sep 23, 2015

initial upgrade to cudnnv3, fwd only
yjxiong
committed

Commits on Sep 24, 2015

supported cudnn_v3
yjxiong
committed
Merge branch 'cudnn3' into action_recog
yjxiong
committed
Update README.md
yjxiong
committed

Commits on Sep 25, 2015

reduce cuDNN mem consumption in Testing.
yjxiong
committed
Merge remote-tracking branch 'origin/action_recog' into action_recog
yjxiong
committed

Commits on Sep 30, 2015

Commits on Oct 5, 2015

refine crop/scale code
yjxiong
committed

Commits on Oct 7, 2015

add scatter layer and related logics
yjxiong
committed

Commits on Oct 8, 2015

add more cropping options
yjxiong
committed

Commits on Oct 9, 2015

try to fix cudnn buffer allocation
yjxiong
committed

Commits on Oct 10, 2015

Merge branch 'action_recog' into data_parallel
yjxiong
committed
refactored cudnn wrapper.
yjxiong
committed

Commits on Oct 13, 2015

Update README.md
yjxiong
committed

Commits on Oct 16, 2015

add normalize, smoothL1Loss layers
yjxiong
committed
Merge remote-tracking branch 'origin/action_recog' into action_recog
yjxiong
committed

Commits on Oct 19, 2015

add frozen state to bn for finetuning on small dataset.
yjxiong
committed

Commits on Oct 20, 2015

expose phase to python layer
yjxiong
committed

Commits on Oct 22, 2015

force releasing opencv mat
yjxiong
committed
minor change
yjxiong
committed

Commits on Nov 11, 2015

add python prefetch thread registry
yjxiong
committed
finish prefetch cpp side
yjxiong
committed

Commits on Nov 13, 2015

Commits on Nov 14, 2015

python layer prefetch tested
yjxiong
committed

Commits on Nov 19, 2015

Merge branch 'action_recog' of https://github.com/yjxiong/caffe into action_recog
yjxiong
committed
remove redundant logging
yjxiong
committed
remove faulty include
yjxiong
committed

Commits on Nov 23, 2015

Update README.md
yjxiong
committed

Commits on Dec 1, 2015

fix testing bug when crop_size not specified
yjxiong
committed

Commits on Dec 15, 2015

Fix CuDNNConvolutionLayer for cuDNN v4

flx42
authored and
yjxiong
committed
Merge branch 'action_recog' of https://github.com/yjxiong/caffe into action_recog
yjxiong
committed

Commits on Dec 17, 2015

Commits on Dec 20, 2015

Find the best cudnn conv algorithm during runtime
Cysu
committed

Commits on Dec 21, 2015

Commits on Dec 22, 2015

Commits on Dec 23, 2015

minor fix for stable optimiz
yjxiong
committed
change mem_tick design
yjxiong
committed
fix zero richness
yjxiong
committed
fix zero richness
yjxiong
committed
change default richness setting
yjxiong
committed
Merge pull request #19 from yjxiong/runtime_optim_cudnn
yjxiong
committed
Minor change
yjxiong
committed
Update README.md
yjxiong
committed
Update CHANGELOG.md
yjxiong
committed
Remove earlier updates
yjxiong
committed
Update README.md
yjxiong
committed
Update README.md
yjxiong
committed
Update CHANGELOG.md
yjxiong
committed
correction of cudnn_conv_layer.cpp
wanglimin
committed
enable travis ci
Yuanjun Xiong
committed
enable travis ci
Yuanjun Xiong
committed
fix travis build
yjxiong
committed
fix travis
yjxiong
committed
fix travis compile failure
yjxiong
committed
disable lint on travis
yjxiong
committed
cannot run make runtime on travis with CUDA
yjxiong
committed

Commits on Dec 24, 2015

Commits on Dec 30, 2015

Commits on Jan 5, 2016

Commits on Jan 7, 2016

Python wrappers for temporal/spatial prediction
Chuck Cho
committed

Commits on Jan 8, 2016

added keyword arguments
Chuck Cho
committed

Commits on Jan 14, 2016

Commits on Jan 16, 2016

Commits on Jan 18, 2016

Update README.md
yjxiong
committed

Commits on Feb 2, 2016

fallback to caffe's BN engine for accuracy issue.
yjxiong
committed

Commits on Feb 17, 2016

Commits on Feb 19, 2016

Commits on Feb 24, 2016

Sync loss when test
Cysu
committed

Commits on Feb 25, 2016

Merge pull request #48 from Cysu/fix-test-sync-loss
yjxiong
committed

Commits on Mar 5, 2016

add scale and bias layers
yjxiong
committed
add scale and bias layers
yjxiong
committed

Commits on Mar 7, 2016

Update video_data_layer.cpp
wanglimin
committed

Commits on Mar 8, 2016

add batch red layer
yjxiong
committed
bread tested
Yuanjun Xiong
committed

Commits on Mar 9, 2016

add gpu implementation
yjxiong
committed

Commits on Mar 10, 2016

set zero
Yuanjun Xiong
committed

Commits on Mar 21, 2016

Merge branch 'action_recog' of https://github.com/yjxiong/caffe into bread
yjxiong
committed

Commits on Mar 30, 2016

add custom file patterns for video data.
yjxiong
committed
Update travis_install.sh
yjxiong
committed
Update travis_install.sh
yjxiong
committed
fix cpu build
yjxiong
committed

Commits on Apr 14, 2016

new reduction options.
yjxiong
committed

Commits on Apr 15, 2016

fix bugs when the video has only frame
yjxiong
committed

Commits on Apr 25, 2016

init cudnn_v5 support
yjxiong
committed

Commits on Apr 26, 2016

Commits on Apr 27, 2016

update bn_convert script
yjxiong
committed
typo
yjxiong
committed
typo
yjxiong
committed
add epsilon option
yjxiong
committed
enable google-style data augmentation.
yjxiong
committed
Cudnn v5 Support (#71)
yjxiong
committed

Commits on Apr 28, 2016

Commits on May 10, 2016

init mem optimizer
yjxiong
committed
fix share diff cases.
yjxiong
committed
share conv buffer and use fastest mode
yjxiong
committed

Commits on May 17, 2016

Merge remote-tracking branch 'origin/bread' into new_anet
yjxiong
committed
fix topk bug
yjxiong
committed

Commits on May 18, 2016

fix snapshot and mem opt time
yjxiong
committed

Commits on May 20, 2016

Update solver.cpp to fix the duplicate snapshot (#78)

Cysu
authored and
yjxiong
committed
Merge remote-tracking branch 'origin/action_recog' into anet
yjxiong
committed
improve mem
yjxiong
committed

Commits on Jun 10, 2016

Update README.md
yjxiong
authored

Commits on Jun 15, 2016

Improve the python interface (#80)

Cysu
authored and
yjxiong
committed

Commits on Jul 16, 2016

Update README.md
yjxiong
authored

Commits on Jul 20, 2016

Reuse data memory if not propagate down (#86)

Cysu
authored and
yjxiong
committed
readme and comments
yjxiong
committed
Merge remote-tracking branch 'origin/org_size_processing' into mem
yjxiong
committed
more comments
yjxiong
committed

Commits on Jul 29, 2016

Update README.md (#95)

ae86208
authored and
yjxiong
committed

Commits on Jul 30, 2016

Commits on Aug 4, 2016

Commits on Aug 11, 2016

standardize memory optimization configurations (#99)
yjxiong
authored

Commits on Sep 2, 2016

Split MPI comm by specified groups (#101)

Cysu
authored and
yjxiong
committed

Commits on Aug 19, 2017

untrimmednet push
wanglimin
committed