Soft-to-Hard Vector Quantization for End-to-End Learned Compression of Images and Neural Networks.
Eirikur AgustssonFabian MentzerMichael TschannenLukas CavigelliRadu TimofteLuca BeniniLuc Van GoolPublished in: CoRR (2017)
Keyphrases
- end to end
- vector quantization
- image compression
- vector quantized
- distortion measure
- neural network
- text localization and recognition
- vector quantizer
- image data
- codebook design
- compression ratio
- compression algorithm
- congestion control
- compression scheme
- transform coding
- image classification
- subband
- low bit rate
- scalable video
- wavelet transform
- image processing
- computer vision