Back Transcription as a Method for Evaluating Robustness of Natural Language Understanding Models to Speech Recognition Errors

Marek Kubis, Paweł Skórzewski, Marcin Sowański, Tomasz Ziętkiewicz
10/25/2023
cs.CLcs.AIcs.SD

No official code found. Best implementation: marekkubis/bteval (3 stars, MIT, last commit 2024-02-17).

Abstract

In a spoken dialogue system, an NLU model is preceded by a speech recognition system that can deteriorate the performance of natural language understanding. This paper proposes a method for investigating the impact of speech recognition errors on the performance of natural language understanding models. The proposed method combines the back transcription procedure with a fine-grained technique for categorizing the errors that affect the performance of NLU models. The method relies on the usage of synthesized speech for NLU evaluation. We show that the use of synthesized speech in place of audio recording does not change the outcomes of the presented technique in a significant way.

View on arXivView PDF

Code Implementations(1)

marekkubis/btevalCommunity100%
MIT
30PythonFeb 3, 20242 years agoMIT

Similar Papers

Exploring the Integration of Large Language Models into Automatic Speech Recognition Systems: An Empirical Study

83% match2023

Modality Confidence Aware Training for Robust End-to-End Spoken Language Understanding

83% match2023

Multimodal Audio-textual Architecture for Robust Spoken Language Understanding

83% match2023

Optimized Tokenization for Transcribed Error Correction

82% match2023

AgenticASR: Refining Speech Recognition in Real-World Scenarios via an Agentic Approach

81% match1 repo12026

Adapting an ASR Foundation Model for Spoken Language Assessment

81% match2023

Cite this paper

@article{kubis2023back,
  title  = {Back Transcription as a Method for Evaluating Robustness of Natural Language Understanding Models to Speech Recognition Errors},
  author = {Marek Kubis and Paweł Skórzewski and Marcin Sowański and Tomasz Ziętkiewicz},
  year   = {2023},
  eprint = {2310.16609},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2310.16609v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2310.16609.svg)](https://codeofpaper.com/paper/2310.16609)

Discussion