N-best T5: Robust ASR Error Correction using Multiple Input Hypotheses and Constrained Decoding Space

Abstract

Error correction models form an important part of Automatic Speech Recognition (ASR) post-processing to improve the readability and quality of transcriptions. Most prior works use the 1-best ASR hypothesis as input and therefore can only perform correction by leveraging the context within one sentence. In this work, we propose a novel N-best T5 model for this task, which is fine-tuned from a T5 model and utilizes ASR N-best lists as model input. By transferring knowledge from the pre-trained language model and obtaining richer information from the ASR decoding space, the proposed approach outperforms a strong Conformer-Transducer baseline. Another issue with standard error correction is that the generation process is not well-guided. To address this a constrained decoding process, either based on the N-best list or an ASR lattice, is used which allows additional information to be propagated.

Proceedings of INTERSPEECH

References in corpus (3)

• Robust Speech Recognition via Large-Scale Weak Supervision
• Error Correction in ASR using Sequence-to-Sequence Models
• Automatic Spelling Correction with Transformer for CTC-based End-to-End Speech Recognition

Cited by in corpus (6)

• Generative Speech Recognition Error Correction with Large Language Models and Task-Activating Prompting
• Whispering LLaMA: A Cross-Modal Generative Error Correction Framework for Speech Recognition
• Large Language Model Based Generative Error Correction: A Challenge and Baselines for Speech Recognition, Speaker Tagging, and Emotion Recognition
• Adapting an Unadaptable ASR System
• FlanEC: Exploring Flan-T5 for Post-ASR Error Correction
• Game-Oriented ASR Error Correction via RAG-Enhanced LLM