Enhancing the Robustness of Contextual ASR to Varying Biasing Information Volumes Through Purified Semantic Correlation Joint Modeling
Authors:
- Yue GuYGYue Gu
- Zhihao DuZDZhihao Du
- Ying ShiYSYing Shi
- Shiliang ZhangSZShiliang Zhang
- Qian ChenQCQian Chen
- Jiqing HanJHJiqing Han
Preprint
Abstract
Recently, cross-attention-based contextual automatic speech recognition (ASR) models have made notable advancements in recognizing personalized biasing phrases. However, the effectiveness of cross-attention is affected by variations in biasing information volume, especially when the length of the biasing list increases significantly. We find that, regardless of the length of the biasing list, only a limited amount of biasing information is most relevant to a specific ASR intermediate representation. Therefore, by identifying and integrating the most relevant biasing information rather than the entire biasing list, we can alleviate the effects of variations in biasing information volume for contextual ASR. To this end, we propose a purified semantic correlation joint modeling (PSC-Joint) approach. In PSC-Joint, we define and calculate three semantic correlations between the ASR intermediate representations and biasing information from coarse to fine: list-level, phrase-level, and token-level. Then, the three correlations are jointly modeled to produce their intersection, so that the most relevant biasing information across various granularities is highlighted and integrated for contextual recognition. In addition, to reduce the computational cost introduced by the joint modeling of three semantic correlations, we also propose a purification mechanism based on a grouped-and-competitive strategy to filter out irrelevant biasing phrases. Compared with baselines, our PSC-Joint approach achieves average relative F1 score improvements of up to 21.34% on AISHELL-1 and 28.46% on KeSpeech, across biasing lists of varying lengths.Keywords
All Related Versions
This publication has 56 references indexed in Scilit:
- Locality Enhanced Dynamic Biasing and Sampling Strategies For Contextual ASR2023
- Context-Aware Transformer Transducer for Speech Recognition2021
- Tree-Constrained Pointer Generator for End-to-End Contextual Speech Recognition2021
- RNN-T Based Open-Vocabulary Keyword Spotting in Mandarin with Multi-Level Detection2021
- End-to-end Contextual Speech Recognition Using Class Language Models and a Token Passing Decoder2019
- Streaming End-to-end Speech Recognition for Mobile Devices2019
- Contextual Speech Recognition with Difficult Negative Training Examples2019
- Minimum Word Error Rate Training for Attention-Based Sequence-to-Sequence Models2018
- Listen, attend and spell: A neural network for large vocabulary conversational speech recognition2016
- Speech recognition with deep recurrent neural networks2013