This paper studies contextual biasing with Large Language Models (LLMs), where during second-pass rescoring additional contextual information is provided to a LLM to boost Automatic Speech Recognition (ASR) performance. We propose to leverage prompts for a LLM without fine tuning during rescoring which incorporate a biasing list and few-shot examples to serve as additional information when calculating the score for the hypothesis. In addition to few-shot prompt learning, we propose multi-task training of the LLM to predict both the entity class and the next token. To improve the efficiency for contextual biasing and to avoid exceeding LLMs' maximum sequence lengths, we propose dynamic prompting, where we select the most likely class using the class tag prediction, and only use entities in this class as contexts for next token prediction. Word Error Rate (WER) evaluation is performed on i) an internal calling, messaging, and dictation dataset, and ii) the SLUE-Voxpopuli dataset. Results indicate that biasing lists and few-shot examples can achieve 17.8% and 9.6% relative improvement compared to first pass ASR, and that multi-task training and dynamic prompting can achieve 20.0% and 11.3% relative WER improvement, respectively.

翻译：本文研究基于大语言模型（LLM）的上下文偏向技术，即在二次解码重评分阶段，通过向大语言模型提供额外上下文信息来提升自动语音识别（ASR）性能。我们提出在重评分过程中无需微调即可利用大语言模型的提示（prompt），通过融合偏向列表（biasing list）和少样本示例（few-shot examples）作为假设评分时的补充信息。除少样本提示学习外，我们还提出对大语言模型进行多任务训练，使其同时预测实体类别和下一个词元。为提升上下文偏向效率并避免超过大语言模型的最大序列长度限制，我们提出动态提示机制：基于类别标签预测选择最可能的类别，仅将该类别中的实体作为下一词元预测的上下文。在i)内部通话、消息和听写数据集及ii)SLUE-Voxpopuli数据集上进行词错误率（WER）评估。结果表明，与首次解码ASR相比，偏向列表和少样本示例分别能实现17.8%和9.6%的相对性能提升，而多任务训练和动态提示机制则分别实现20.0%和11.3%的相对WER改善。