A regularizer that maximizes conditional source entropy gives BD-rate gains of 0.9% to 3.0% across five neural compression models, but may be equivalent to reweighting the rate term.
An introduction to neural data compression,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
An Information-Theoretic Regularizer for Lossy Neural Image Compression
A regularizer that maximizes conditional source entropy gives BD-rate gains of 0.9% to 3.0% across five neural compression models, but may be equivalent to reweighting the rate term.