Proposes deterministic LBRC and neural NGLR residual coders that improve compression ratios 30-100% over GAE for block-level NRMSE targets of 10^-6 to 10^-4 on E3SM, JHTDB, and ERA5 data.
LPCNet: Improving neural speech synthe- sis through linear prediction,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Residual Modeling for High-Fidelity Learned Compression of Scientific Data
Proposes deterministic LBRC and neural NGLR residual coders that improve compression ratios 30-100% over GAE for block-level NRMSE targets of 10^-6 to 10^-4 on E3SM, JHTDB, and ERA5 data.