A Large RGB-D Gait Dataset and the Baseline Algorithm

With the development of depth sensors, images with high quality depth can be obtained easily. Using depth information, some challenging problems in gait recognition can be reconsidered and better solutions can be developed. To prompt gait recognition with depth information, a large RGB-D gait dataset is introduced. It contains 99 subjects, with 8 sequences for each subjects in two different views. A baseline algorithm, namely Gait Energy Surface (GES), is proposed for researchers to evaluate their own algorithms. Even it is a baseline algorithm, encouraging experimental results have been achieved.