ViCo-MoCo-DL: Video Coding and Motion Compensation Solutions for Human Activity Recognition Using Deep Learning.
Tamer ShanablehPublished in: IEEE Access (2023)
Keyphrases
- motion compensation
- video coding
- deep learning
- human activities
- motion compensated
- video compression
- motion estimation
- motion vectors
- bit rate
- rate distortion
- activity recognition
- motion estimation and compensation
- motion field
- unsupervised learning
- video quality
- motion estimation algorithm
- motion compensated prediction
- video coder
- block matching
- machine learning
- human computer interaction
- error resilience
- video codec
- video coding standard
- rate control
- macroblock
- low bit rate
- compression efficiency
- action recognition
- computational complexity
- image sequences
- high quality
- supervised learning
- motion model
- video data