Replacing STFT with a real-valued singular value decomposition-based graph transform improves speech enhancement across multiple neural backbones.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.AS 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Time-Graph Frequency Representation with Singular Value Decomposition for Neural Speech Enhancement
Replacing STFT with a real-valued singular value decomposition-based graph transform improves speech enhancement across multiple neural backbones.