A two-stage network that decouples deep filtering into temporal and frequency stages achieves near top scores on VoiceBank+DEMAND with far fewer parameters than most systems.
We also introduce the TAConv module, which further captures temporal correlations for speech enhancement tasks
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.SD 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
A Two-Stage Hierarchical Deep Filtering Framework for Real-Time Speech Enhancement
A two-stage network that decouples deep filtering into temporal and frequency stages achieves near top scores on VoiceBank+DEMAND with far fewer parameters than most systems.