MiniViT: Compressing Vision Transformers with Weight Multiplexing.
Jinnian ZhangHouwen PengKan WuMengchen LiuBin XiaoJianlong FuLu YuanPublished in: CVPR (2022)
Keyphrases
- computer vision
- vision system
- real time
- data compression
- databases
- weight assignment
- three dimensional
- weighting scheme
- video streams
- visually guided
- database
- computational vision
- weight function
- feature weights
- compressed data
- artificial neural networks
- computational complexity
- data structure
- website
- image processing
- learning algorithm
- neural network