Encoder-Decoder Based Convolutional Neural Networks with Multi-Scale-Aware Modules for Crowd Counting.
Pongpisit ThanasutivesKen-ichi FukuiMasayuki NumaoBoonserm KijsirikulPublished in: ICPR (2020)
Keyphrases
- convolutional neural networks
- multiscale
- decoding process
- video codec
- distributed video coding
- error control
- low complexity
- wyner ziv
- noisy channel
- convolutional network
- successive approximation
- rate distortion
- wyner ziv video coding
- video coding
- mpeg avc
- distributed source coding
- video coding scheme
- motion compensated prediction
- error correction
- transform domain
- motion estimation
- pixel domain
- turbo codes
- scale space
- temporal correlation
- bit rate
- rate allocation
- error resilience
- edge detection
- low bit rate
- inter frame
- bitstream
- compressive sensing
- motion compensated
- macroblock
- vector quantization
- coarse to fine
- video quality
- video encoder
- decoding algorithm
- compressed video
- error concealment
- rate control
- motion compensation
- bit budget
- image segmentation
- video sequences