Robust RGB-D Face Recognition using Kinect Sensor

Robust RGB-D Face Recognition using Kinect Sensor

时间： 地点：九龙湖校区计算机楼四楼会议室

报告简介：

We propose a robust face recognition algorithm for low resolution RGB-D Kinect data. During preprocessing, facial symmetry is exploited at the 3D point cloud level to obtain a canonical frontal view irrespective of the initial pose. Depth data is converted to XYZ normal maps. Multichannel Discriminant Transforms are then used to project RGB to DCS (Discriminant Color Space) and normal maps to DNM (Discriminant Normal Maps). A Multi-channel Robust Sparse Coding method is proposed that codes the multiple channels (DCS or DNM) of a test image as a sparse combination of training samples with different pixel weighting. Weights are calculated dynamically in an iterative process to achieve robustness to variations in pose, illumination, facial expressions and disguise. In contrast to existing techniques, our multi-channel approach is more robust to variations. Reconstruction errors of the test image (DCS and DNM) are normalized and fused to decide its identity. The proposed algorithm was evaluated on four public databases. We achieved 98.4% identification rate on CurtinFaces, a Kinect database with 4784 RGB-D images of 52 subjects. Using a first versus all protocol on the Bosphorus, CASIA and FRGC v2 databases, we achieved 97.6%, 95.6% and 95.2% identification rates respectively. To the best of our knowledge, these are the highest identification rates reported to date for the first three databases.

报告人简介：

Dr. Wanquan Liu received the BSc degree in Applied Mathematics from Qufu Normal University, P. R. China, in 1985, the MSc degree in Control Theory and Operation Research from Chinese Academy of Science in 1988, and the PhD degree in Electrical Engineering from Shanghai Jiaotong University, in 1993. He once held the ARC Fellowship, U2000 Fellowship and JSPS Fellowship and attracted research funds from different resources over 2 million dollars. He is currently an Associate Professor in the Department of Computing at Curtin University and is in editorial board for ten international journals. His current research interests include large-scale pattern recognition, signal processing, machine learning, and control systems. His publications have been cited over 2500 times from Google scholar.

计算机网络和信息集成教育部重点实验室（东南大学）