The dataset was captured by a Kinect device. There are 12 dynamic American Sign Language (ASL) gestures, and 10 people. Each person performs each gesture 2-3 times. There are 336 files in total, each corresponding to a depth sequence. The hand portion (above the wrist) has been segmented. The file name has the format sub_depth_m_n where m is the person index. n ranges from 1 to 36. Note that for some (m,n), the file sub_depth_m_n does not exist. For example, there is no "sub_depth_02_03". The reason is that some of the bad sequences are excluded from the dataset. The mapping from n to gesture type is the following:
{1,2,3}-> "ASL_Z";
{4,5,6} ->"ASL_J";
{7,8,9} ->"ASL_Where";
{10,11,12} ->"ASL_Store";
{13,14,15} ->"ASL_Pig";
{16,17,18} ->"ASL_Past";
{19,20,21}->"ASL_Hungary";
{22.23,24}->"ASL_Green";
{25.26.27}->"ASL_Finish";
{28,29,30}->"ASL_Blue";
{31,32,33}->"ASL_Bathroom";
{34,35,36}->"ASL_Milk";
Each file is a MAT file which can be loaded with 64bit MATLAB. Below is a sample MATLAB code to load a file:
x=load('sub_depth_01_01');
width = size(x.depth_part,1);
height = size(x.depth_part,2);
nFrames = size(x.depth_part,3);
for(i=1:width)
for(j=1:height)
for(k=1:nFrames)
depthval = x.depth_part(i,j,k);
end
end
end
The following two papers reported experiment results on this dataset:
Alexey Kurakin, Zhengyou Zhang, Zicheng Liu, A Real-Time System for Dynamic Hand Gesture Recognition with a Depth Sensor, EUSIPCO, 2012.
Jiang Wang, Zicheng Liu, Jan Chorowski, Zhuoyuan Chen, Ying Wu, Robust 3D Action Recognition with Random Occupancy Patterns, ECCV, 2012.