AINeutralarXiv – CS AI · Mar 54/10
🧠
ACES: Accent Subspaces for Coupling, Explanations, and Stress-Testing in Automatic Speech Recognition
Researchers introduce ACES, a new method to analyze how automatic speech recognition systems perform differently across accents. The study finds that accent information is concentrated in early neural network layers and is deeply intertwined with speech recognition capabilities, making simple bias removal ineffective.