Visual-only Voice Activity Detection using Human Motion in Conference Video.
Keisuke YamazakiSatoshi TamuraYuuto GotohMasaki NosePublished in: ICPRAM (2022)
Keyphrases
- human motion
- visual data
- video sequences
- human actions
- voice activity detection
- motion capture data
- articulated human motion
- motion capture
- human body
- motion trajectories
- video data
- image sequences
- temporal segmentation
- optical motion capture
- gait recognition
- motion sequences
- video frames
- motion recognition
- human motion tracking
- motion history images
- space time
- human movements
- visual information
- spatio temporal
- news video
- video content
- body parts
- motion synthesis
- video shots
- moving objects
- human motion analysis
- noisy environments
- low level
- motion estimation
- visual features
- key frames
- multi view
- video analysis
- camera motion
- spatial and temporal
- three dimensional
- multimedia
- machine learning