1 paper
Pratik Rakesh Singh, Mohammadi Zaki, Aneesh Mukkamala +1
Automatic speech recognition (ASR) systems, despite low overall word error rates, produce residual lexical errors that disproportionately affect semantically critical tokens such a…