DMM.zip_kinect_kinect 人体_人体 动作_深度图像_深度图像 matlab

******************************************************************************* Demo Codes For Action Recognition Based on Depth Motion Maps by Chen Chen, The University of Texas at Dallas ******************************************************************************* Reference C. Chen, K. Liu, and N. Kehtarnavaz, "Real-Time Human Action Recognition Based on Depth Motion Maps," Journal of Real-Time Image Processing, to appear. Please cite our work if you use the code. For any problems, please send email to chenchen870713@gmail.com Thanks! =============================================================================== 1. Test dataset: Microsoft Research (MSR) Action3D, which can be found here: http://research.microsoft.com/en-us/um/people/zliu/actionrecorsrc/ 20 action types, 10 subjects, each subject performs each action 2 or 3 times. There are 567 depth map sequences in total. The resolution is 320x240. The depth sequences for each action are stored as MATLAB mat file, which is a matrix of size 320x240x(number of depth frames). 2. Test_1And2_Fix.m: Samples for training and testing are chosen as FIXED order for test one and two. 3. Test_1And2_Random.m: Samples for training and testing are chosen as RANDOM for test one and two. 4. Test_Cross_Fix.m: Subjects for training and testing are FIXED for cross subject test. Subject 1 3 5 7 9 as training subjects, the rest as testing subjects. 5. Test_Cross_Random.m: Subjects for training and testing are chosen as RANDOM.