Fast-ULCNet matches original ULCNet speech enhancement quality while cutting model size by more than half and latency by 34% via FastGRNN replacement and a state-drift filter.
Build- ing upon the low-complexity, state-of-the-art ULCNet archi- tecture, we propose replacing its GRU layers with FastGRNN units to further reduce computational overhead
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
-
Fast-ULCNet: A fast and ultra low complexity network for single-channel speech enhancement
Fast-ULCNet matches original ULCNet speech enhancement quality while cutting model size by more than half and latency by 34% via FastGRNN replacement and a state-drift filter.