CCV Database Release 
---------------------------------------------------------------------
categoryName: names of 20 CCV categories

trainVidID: unique YouTube IDs of 4659 training videos

testVidID: unique YouTube IDs of 4658 test videos

trainLabel: 4659x20 label matrix (rows and columns follow video/category orders in trainVidID and categoryName, respectively)

testLabel: 4658x20 label matrix (rows and columns follow video/category orders in testVidID and categoryName, respectively)

STIP-trainFeature: 4659x5000 STIP feature matrix for training videos

STIP-testFeature: 4658x5000 STIP feature matrix for test videos

SIFT-trainFeature: 4659x5000 SIFT feature matrix for training videos

SIFT-testFeature: 4658x5000 SIFT feature matrix for test videos

MFCC-trainFeature: 4659x4000 MFCC feature matrix for training videos

MFCC-testFeature: 4658x4000 MFCC feature matrix for test videos
----------------------------------------------------------------------

Citation:
Yu-Gang Jiang, Guangnan Ye, Shih-Fu Chang, Daniel Ellis, Alexander C. Loui, Consumer Video Understanding: A Benchmark Database and An Evaluation of Human and Machine Performance, ACM International Conference on Multimedia Retrieval (ICMR), Trento, Italy, Apr. 2011.


--
Yu-Gang Jiang (yjiang@ee.columbia.edu)
Columbia University
4/8/2011