Cache Compression with Golomb-Rice Code and Quantization for Convolutional Neural Networks.
Seung-Hwan BaeHyuk-Jae LeeHyun KimPublished in: ISCAS (2021)
Keyphrases
- convolutional neural networks
- entropy coding
- lossless compression
- compression scheme
- image compression
- arithmetic coding
- huffman coding
- lossy image compression
- convolutional network
- compression ratio
- uniform quantization
- data compression
- quantization noise
- progressive transmission
- transform coding
- efficient compression
- compression algorithm
- low complexity
- prefetching
- code length
- processor core
- bit rate
- source code
- query processing
- data access
- quantization scheme
- variable length
- bits per pixel
- gray code
- coding method
- reconstructed image
- computational complexity
- vector quantization
- inter frame
- adaptive quantization
- block size
- quantization error
- linear prediction
- block coding
- industry standard
- main memory
- response time
- transform coefficients