Contextualization of ASR with LLM using phonetic retrieval-based augmentation
Authors:
- 6 April 2025
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Abstract
Large language models (LLMs) have shown superb capability of modeling multimodal signals including audio and text, allowing the model to generate spoken or textual response given a speech input. However, it remains a challenge for the model to recognize personal named entities, such as contacts in a phone book, when the input modality is speech. In this work, we start with a speech recognition task and propose a retrievalbased solution to contextualize the LLM: we first let the LLM detect named entities in speech without any context, then use this named entity as a query to retrieve phonetically similar named entities from a personal database and feed them to the LLM, and finally run context-aware LLM decoding. In a voice assistant task, our solution achieved up to 30.2% relative word error rate reduction and 73.6% relative named entity error rate reduction compared to a baseline system without contextualization. Notably, our solution by design avoids prompting the LLM with the full named entity database, making it highly efficient and applicable to large named entity databases.Keywords
All Related Versions
This publication has 13 references indexed in Scilit:
- Connecting Speech Encoder and Large Language Model for ASR2024
- Personalization of CTC-Based End-to-End Speech Recognition Using Pronunciation-Driven Subword Tokenization2024
- Prompting Large Language Models with Speech Recognition Abilities2024
- End-to-End Speech Recognition Contextualization with Large Language Models2024
- SALM: Speech-Augmented Language Model with in-Context Learning for Speech Recognition and Translation2024
- Integrating Pre-Trained Speech and Language Models for End-to-End Speech Recognition2024
- Acoustic Model Fusion For End-to-End Speech Recognition2023
- Revisiting Large Language Models as Zero-shot Relation Extractors2023
- Deep Context: End-to-end Contextual Speech Recognition2018
- Connectionist temporal classification2006