Customizing Speech Recognition Model with Large Language Model Feedback

Shaoshi Ling, Guoli Ye

Automatic speech recognition (ASR) systems have achieved strong performance on general transcription tasks. However, they continue to struggle with recognizing rare named entities…