PROCTER: PROnunciation-aware ConTextual adaptER for personalized speech recognition in neural transducers
CoRR(2023)
摘要
End-to-End (E2E) automatic speech recognition (ASR) systems used in voice
assistants often have difficulties recognizing infrequent words personalized to
the user, such as names and places. Rare words often have non-trivial
pronunciations, and in such cases, human knowledge in the form of a
pronunciation lexicon can be useful. We propose a PROnunCiation-aware
conTextual adaptER (PROCTER) that dynamically injects lexicon knowledge into an
RNN-T model by adding a phonemic embedding along with a textual embedding. The
experimental results show that the proposed PROCTER architecture outperforms
the baseline RNN-T model by improving the word error rate (WER) by 44
when measured on personalized entities and personalized rare entities,
respectively, while increasing the model size (number of trainable parameters)
by only 1
personalized device names, we observe 7
compared to only 1
更多查看译文
关键词
personalized speech
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要