Naijun Zheng, Xucheng Wan, Kai Liu, Ziqing Du, Zhou Huan
TL;DR使用简单的文本增强技术借助大量纯文本数据集来构建编码簿,可以提高预训练的 ASR 模型的上下文信息,从而显著提升识别性能。
Abstract
Although contextualized automatic speech recognition (ASR) systems are
commonly used to improve the recognition of uncommon words, their effectiveness
is hindered by the inherent limitations of speech-text data availability. To
address this challenge, our study proposes to leverage ext