Exploiting temporal and depth information for multi-frame face anti-spoofing.
Zezheng WangChenxu ZhaoYunxiao QinQiusheng ZhouZhen LeiPublished in: CoRR (2018)
Keyphrases
- depth information
- multi frame
- temporal coherence
- depth map
- motion estimation
- stereo vision
- image sequences
- point correspondences
- super resolution
- depth recovery
- computationally efficient
- optical flow
- stereo camera
- depth images
- motion analysis
- filtering algorithm
- optic flow
- stereo correspondence
- video frames
- multi view
- vision system
- real time
- multiscale
- disparity map
- image quality
- stereo matching
- motion model
- low resolution
- single image