Search Results for author:
Found 2 papers, 1 papers with code
WHAMR!: Noisy and Reverberant Single-Channel Speech Separation
While significant advances have been made with respect to the separation of overlapping speech signals, studies have been largely constrained to mixtures of clean, near anechoic speech, not representative of many real-world scenarios.
Sound Audio and Speech Processing
WHAM!: Extending Speech Separation to Noisy Environments
Recent progress in separating the speech signals from multiple overlapping speakers using a single audio channel has brought us closer to solving the cocktail party problem.
Ranked #15 on
Speech Separation
on WHAMR!
