文件列表:

classification.m (622, 2014-09-25)
compute_feature.m (536, 2014-09-25)
fuse_training.m (438, 2014-09-25)
get_global.m (902, 2014-09-25)
initialization4LCKSVD.m (1723, 2014-09-25)
labelconsistentDTLC.m (718, 2014-09-25)
main.m (5014, 2014-09-26)
rand_sampling.m (650, 2009-09-27)
retr_database_dir.m (1086, 2014-09-26)
retrieve_dictionary.m (548, 2014-09-25)
sc_approx_pooling.m (2610, 2012-06-17)
sc_pooling.m (2451, 2012-06-17)
dictionary\dict_img_depth_512.mat (20651, 2014-09-25)
ksvdbox\Contents.m (1068, 2009-08-03)
ksvdbox\faq.txt (4348, 2009-10-18)
ksvdbox\ksvd.m (19242, 2013-09-26)
ksvdbox\ksvddemo.m (1736, 2009-07-27)
ksvdbox\ksvddenoise.m (10724, 2009-10-03)
ksvdbox\ksvddenoisedemo.m (2415, 2009-08-24)
ksvdbox\ksvdver.m (13711, 2009-10-18)
ksvdbox\odct2dict.m (1057, 2009-07-27)
ksvdbox\odct3dict.m (1149, 2009-07-27)
ksvdbox\odctdict.m (452, 2009-07-27)
ksvdbox\odctndict.m (1459, 2009-07-27)
ksvdbox\ompdenoise.m (10423, 2009-10-03)
ksvdbox\ompdenoise1.m (4084, 2009-09-07)
ksvdbox\ompdenoise2.m (4059, 2009-09-07)
ksvdbox\ompdenoise3.m (4229, 2009-09-07)
ksvdbox\private\addtocols.c (1915, 2009-07-27)
ksvdbox\private\addtocols.m (303, 2009-07-27)
ksvdbox\private\addtocols.mexw64 (8192, 2010-09-14)
ksvdbox\private\add_dc.m (830, 2009-07-27)
ksvdbox\private\col2imstep.c (3924, 2009-08-31)
ksvdbox\private\col2imstep.m (1001, 2009-08-31)
ksvdbox\private\col2imstep.mexw64 (9728, 2010-09-14)
ksvdbox\private\collincomb.c (4255, 2009-07-27)
ksvdbox\private\collincomb.m (832, 2009-07-27)
ksvdbox\private\collincomb.mexw64 (10240, 2010-09-14)
ksvdbox\private\countcover.m (1178, 2009-09-01)
... ...

Relevant techniques utilized in this implementation: ScSPM (http://www.ifp.illinois.edu/~jyang29/ScSPM.htm) K-SVD (http://www.cs.technion.ac.il/~ronrubin/software.html) LC-KSVD (http://www.umiacs.umd.edu/~zhuolin/projectlcksvd.html) All codes are provided for noncommercial research use If you happen to use this code, please cite our work: [1] F. Zhu and L. Shao, °Weakly-Supervised Cross-Domain Dictionary Learning for Visual Recognition±, International Journal of Computer Vision (IJCV), vol. 109, no. 1-2, pp. 42-59, Aug. 2014. [2] F. Zhu and L. Shao, °Enhancing Action Recognition by Cross-Domain Dictionary Learning±, British Machine Vision Conference (BMVC), Bristol, UK, 2013. If you find any bugs in this code, please contact (fan.zhu@sheffield.ac.uk) Installation: 1. Pre-compiled mex functions are provided for ***-bit windows systems. If you are using other types of OSs, please compile C files in the private folder of OMPbox and ksvdbox first. 2. If you have not done so before, configure Matlab's MEX compiler by entering >> mex -setup and follow the instructions. Usage: 1. For a quick start, you should copy your data folders (both target domain and source domain) into the ``image'' folder, where the ``image'' locates under the root directory. The standard format of your data folder should look as follows: <-- <-- 2. Set para.dataSet in line 22 of main.m file as your target domain folder_name, and set para.dataSet in line 59 of main.m file as your source folder_name. 3. Annotations of parameter settings are provided in the main.m file. 4. If you are working on video data, or would like to use other image features, you'll have to massage this code to fit your features.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：