DctViT: Discrete Cosine Transform meet vision transformers.
Keke SuLihua CaoBotong ZhaoNing LiDi WuXiyu HanYangfan LiuPublished in: Neural Networks (2024)
Keyphrases
- discrete cosine transform
- image compression
- dct coefficients
- transform domain
- computer vision
- filter bank
- dct domain
- coded images
- discrete fourier transform
- jpeg images
- lapped transforms
- compressed images
- image processing
- blocking artifacts
- image blocks
- spatial domain
- image coding
- super resolution
- similarity measure
- multimedia
- projections onto convex sets