[Demo]Speech Enhancement with Flow Matching Method

 

 

Speech Enhancement with Flow Matching Method

Code here

Train set: WSJ0-C0, Test set: WSJ0-C0

clean noisy Proposed Flow matching NFE=5 SGMSE_CRP NFE = 5 [1] (reimplemented)
440c020a.wav K 218 K 228 K 223 K 233
441o030v.wav K 217 K 227 K 222 K 232
443o0313.wav K 216 K 226 K 221 K 231
445o0303.wav K 215 K 225 K 220 K 230
447o030s.wav K 214 K 224 K 219 K 229

[1] Bunlong Lay, Jean-Marie Lermercier, Julius Richter and Timo Gerkmann. Single and Few-step Diffusion for Generative Speech Enhancement, ICASSP, 2024.

 

Leave a Comment