Plug-and-Play video reconstruction using sparse 3D transform-domain block matching.
Vahid Khorasani GhassabNizar BouguilaPublished in: Mach. Vis. Appl. (2021)
Keyphrases
- transform domain
- block matching
- reconstruction quality
- motion vectors
- motion estimation
- spatial domain
- macroblock
- video sequences
- video data
- compressed sensing
- motion compensation
- video compression
- compressive sensing
- video content
- video coding
- optical flow
- video frames
- dct domain
- block size
- hadamard transform
- motion compensated
- discrete cosine transform
- image sequences
- error concealment
- image blocks
- reference frame
- video codec
- motion field
- prediction error
- low complexity
- inter frame
- frequency domain
- sparse representation
- compressed domain
- image reconstruction
- coding efficiency
- motion model
- computational complexity
- error propagation
- three dimensional
- rate distortion
- dct coefficients
- computer vision
- bit rate
- natural images
- high resolution
- multimedia
- feature selection