S-VVAD: Visual Voice Activity Detection by Motion Segmentation.
Muhammad ShahidCigdem BeyanVittorio MurinoPublished in: WACV (2021)
Keyphrases
- motion segmentation
- voice activity detection
- motion estimation
- missing data
- image sequences
- model selection
- motion model
- moving objects
- multibody
- spectral clustering
- dynamic scenes
- optical flow
- motion sequences
- epipolar geometry estimation
- noisy environments
- feature tracking
- multiple moving objects
- visual features
- motion field
- flow field
- video sequences
- video coding
- pose estimation
- computational complexity
- motion detection
- inter frame
- pattern recognition