Kinect v2 data read and visiualization in tools.

watch_data.m - read single frame from video and visualize it, contains all functions in the tool

depth_plane2depth_world.m - convert depth to depth world pointcould;

get_depth_world_rgb.m - map rgb world coordinates to depth world coordinates;

visualize_human_rgb.m - visualize human body in rgb image;

visualize_human_depth.m - visualize human body in depth image;

visualize_point_cloud.m - visualize colored 3d point cloud;

visualize_joints.m - visualize human joints in 3d point cloud.

Run the training in Catm.

catm = CatmRun(savefid, dataobj, K, W, N, J, T, S)
%K: number of topics
%W: the number of unique words
%N: number of sample rounds
%J: number of LDA initializations
%T: number of times V is sampled
%S: random seed
%dataobj.data.doc: a cell of docs, each of which is a word index vector
%dataobj.data.rtime: a cell of doc's relative time, each of which is nword*nword matrix