Ring mixing and SCER loss break symmetry in noisy speech separation training, allowing models to learn denoising from noisy mixtures alone and halve residual noise on benchmarks.
In all cases, including SCER on mixtures of noisy speech improves SI-SDRi, by 1.2 −1.9 dB, closing about half the gap to the ideal, clean-speech supervision
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.AS 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Ring Mixing with Auxiliary Signal-to-Consistency-Error Ratio Loss for Unsupervised Denoising in Speech Separation
Ring mixing and SCER loss break symmetry in noisy speech separation training, allowing models to learn denoising from noisy mixtures alone and halve residual noise on benchmarks.