It is difficult for an E2E ASR system to recognize words such as entities appearing infrequently in the training data. A widely used method to mitigate this issue is feeding contextual information into the acoustic model. Previous works have proven that a compact and accurate contextual list can boost the performance significantly. In this paper, we propose an efficient approach to obtain a high quality contextual list for a unified streaming/non-streaming based E2E model. Specifically, we make use of the phone-level streaming output to first filter the predefined contextual word list then fuse it into non-casual encoder and decoder to generate the final recognition results. Our approach improve the accuracy of the contextual ASR system and speed up the inference process. Experiments on two datasets demonstrates over 20% CER reduction comparing to the baseline system. Meanwhile, the RTF of our system can be stabilized within 0.15 when the size of the contextual word list grows over 6,000.

翻译：在端到端自动语音识别系统中，识别训练数据中低频出现的词汇（如专有名词）存在困难。缓解该问题的常用方法是将上下文信息输入声学模型。已有研究表明，紧凑且准确的上下文列表可显著提升系统性能。本文针对基于统一流式/非流式的端到端模型，提出了一种获取高质量上下文列表的高效方法。具体而言，我们利用音素级流式输出对预定义的上下文词表进行初步过滤，随后将其融合至非因果编码器和解码器以生成最终识别结果。该方法在提升上下文感知语音识别系统准确率的同时加速了推理过程。在两个数据集上的实验表明，相较于基线系统，字符错误率降低超过20%。此外，当上下文词表规模超过6,000时，系统的实时因子仍能稳定控制在0.15以内。