文件列表:

.dockerignore (14, 2018-08-07)
LICENSE (35141, 2018-08-07)
data (0, 2018-08-07)
data\pose-512.zip (1335239, 2018-08-07)
docker (0, 2018-08-07)
docker\Dockerfile (794, 2018-08-07)
docker\batch_norm.cu (316, 2018-08-07)
docker\mxnet (0, 2018-08-07)
docker\mxnet\Dockerfile (2066, 2018-08-07)
docker\mxnet\dropout-inl.h (8074, 2018-08-07)
docker\mxnet\elementwise_binary_broadcast_op-inl.h (39148, 2018-08-07)
docker\mxnet\resource.cc (8188, 2018-08-07)
docker\mxnet\tensor_gpu-inl.cuh (27183, 2018-08-07)
scripts (0, 2018-08-07)
scripts\app (105, 2018-08-07)
scripts\train_csl.sh (2412, 2018-08-07)
scripts\train_db1.sh (3419, 2018-08-07)
scripts\train_dba.sh (3167, 2018-08-07)
scripts\train_dbb.sh (3190, 2018-08-07)
scripts\train_dbc.sh (3196, 2018-08-07)
scripts\train_table_1.sh (3968, 2018-08-07)
scripts\train_table_4.sh (126, 2018-08-07)
sigr (0, 2018-08-07)
sigr\__init__.py (881, 2018-08-07)
sigr\_app (0, 2018-08-07)
sigr\_app\__init__.py (1448, 2018-08-07)
sigr\_app\app_echo.py (141, 2018-08-07)
sigr\_app\app_test_semimyo.py (12881, 2018-08-07)
sigr\_app\app_train_semimyo.py (6814, 2018-08-07)
sigr\_app\app_train_semimyo_glove.py (7194, 2018-08-07)
sigr\_app\app_train_semimyo_pose_rnn.py (6909, 2018-08-07)
sigr\_app\app_train_semimyo_pose_smooth.py (7245, 2018-08-07)
sigr\_app\app_train_semimyo_pose_stl.py (6916, 2018-08-07)
sigr\_app\app_train_semimyo_v20161213.py (7190, 2018-08-07)
sigr\_app\app_train_semimyo_v20161214.py (6347, 2018-08-07)
sigr\_app\options.py (1113, 2018-08-07)
sigr\_app\train_semimyo.py (10080, 2018-08-07)
... ...

# Semi-Supervised Learning for Surface EMG-based Gesture Recognition ## Prerequisite * A CUDA compatible GPU * Ubuntu 14.04 or any other Linux/Unix that can run Docker * [Docker](http://docker.io/) * [Nvidia Docker](https://github.com/NVIDIA/nvidia-docker) * Download the docker image: ``` docker pull answeror/sigr:semi ``` or build it by yourself: ``` docker build -t answeror/sigr:semi -f docker/Dockerfile . ``` ## Steps to generate table 4 ``` # Prepare data mkdir -p .cache # Download https://www.idiap.ch/project/ninapro # Put NinaPro DB1 in .cache/ninapro-db1-raw # Download http://zju-capg.org/myo/data # Put CapgMyo DB-a in .cache/dba # Put CapgMyo DB-b in .cache/dbb # Put CapgMyo DB-c in .cache/dbc # Download http://www.csl.uni-bremen.de/cms/forschung/bewegungserkennung # Put csl-hdemg in .cache/csl scripts/train_table_4.sh scripts/app test_semimyo --cmd table_4 ``` Training on NinaPro and CapgMyo will take 1 to 2 hours depending on your GPU. Training on csl-hdemg will take several days. You can accelerate traning and testing by distribute different folds on different GPUs with the `gpu` parameter. You can also do train and test for each dataset on different machines or GPUs: ``` scripts/train_db1.sh scripts/train_dba.sh scripts/train_dbb.sh scripts/train_dbc.sh scripts/train_csl.sh scripts/app test_semimyo --cmd table_4_db1 scripts/app test_semimyo --cmd table_4_dba scripts/app test_semimyo --cmd table_4_dbb scripts/app test_semimyo --cmd table_4_dbc scripts/app test_semimyo --cmd table_4_csl ``` ## Steps to generate table 1 ``` # Prepare data mkdir -p .cache # Download https://www.idiap.ch/project/ninapro # Put NinaPro DB1 in .cache/ninapro-db1-raw # Extract pre-calculated cluster (512 clusters) labels to data folder unzip data/pose-512 -d .cache/ninapro-db1-raw scripts/train_table_1.sh scripts/app test_semimyo --cmd table_1 ``` ## License Licensed under an GPL v3.0 license. ## Bibtex ``` @inproceedings{Du_IJCAI_2017, author = {Yu Du, Yongkang Wong, Wenguang Jin, Wentao Wei, Yu Hu, Mohan Kankanhalli, Weidong Geng}, title = {Semi-Supervised Learning for Surface EMG-based Gesture Recognition}, booktitle = {Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence, {IJCAI-17}}, pages = {1624--1630}, year = {2017}, doi = {10.24963/ijcai.2017/225}, url = {https://doi.org/10.24963/ijcai.2017/225}, } ``` ## Misc Thanks DMLC team for their great [MxNet](https://github.com/dmlc/mxnet)!

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：