A Computationally Efficient Neural Video Compression Accelerator Based on a Sparse CNN-Transformer Hybrid Network.
Siyu ZhangWendong MaoHuihong ShiZhongfeng WangPublished in: CoRR (2023)
Keyphrases
- video compression
- computationally efficient
- network architecture
- motion compensation
- motion estimation
- video coding
- motion compensated
- video data
- low bit rate
- video processing
- neural network
- video conferencing
- bit allocation
- video codec
- compression ratio
- block matching
- inter frame
- cellular neural networks
- distribution network
- computational complexity
- multiscale
- image and video compression
- discrete cosine transforms
- image coding
- sparse representation
- multiresolution
- image processing
- computer vision