Multimodal Consistency-Guided Reference-Free Data Selection for ASR Accent Adaptation
arXiv:2602.13263v1 Announce Type: new Abstract: Automatic speech recognition (ASR) systems often degrade on accented speech because acoustic-phonetic and prosodic shifts induce a mismatch to training …
Ligong Lei, Wenwen Lu, Xudong Pang, Zaokere Kadeer, Aishan Wumaier
25 views