Dependent Scalar Quantization For Neural Network Compression.
Paul HaaseHeiko SchwarzHeiner KirchhofferSimon WiedemannTalmaj MarincArturo MarbánKarsten MüllerWojciech SamekDetlev MarpeThomas WiegandPublished in: ICIP (2020)
Keyphrases
- scalar quantization
- image coding
- image compression
- subband
- vector quantization
- coding scheme
- multiple description
- finite state
- source coding
- image coder
- wavelet transform
- rate allocation
- filter bank
- compression algorithm
- multiresolution
- bit rate
- rate distortion optimization
- rate distortion
- compression scheme
- multiple description coding
- information theoretic
- reconstructed image
- compression ratio
- transform coding
- entropy coding
- lossless compression
- vector quantizer
- data compression
- bit plane
- video coding
- coding method
- rate control
- video signals
- multiscale
- end to end
- wavelet coefficients
- markov chain
- feature vectors
- computational complexity