Transform Network Architectures for Deep Learning based End-to-End Image/Video Coding in Subsampled Color Spaces.
Hilmi E. EgilmezAnkitesh K. SinghMuhammed Z. CobanMarta KarczewiczYinhao ZhuYang YangAmir SaidTaco S. CohenPublished in: CoRR (2021)
Keyphrases
- end to end
- video coding
- color space
- rate distortion
- motion compensation
- motion compensated
- bit rate
- input image
- motion estimation
- video quality
- image classification
- video compression
- image data
- image features
- motion vectors
- color images
- image retrieval
- macroblock
- image segmentation
- segmentation method
- error resilience
- color histogram
- test images
- spatial information
- multiscale
- image representation
- spatial domain
- image quality
- object recognition
- computational complexity
- video sequences
- three dimensional