Universally Quantized Neural Compression
Eirikur Agustsson Lucas Theis
Google Research Google Research
eirikur@google.com theis@google.com
Abstract
A popular approach to learning encoders for lossy compression is to use additive
uniform noise during training as a differentiable approximation to test-time quan-
tization. We demonstrate that a uniform noise channel can also b ...


雷达卡


京公网安备 11010802022788号







