What BERT Based Language Models Learn In Spoken Transcripts: An Empirical Study

Abstract

Language Models (LMs) have been ubiquitously leveraged in various tasks including spoken language understanding (SLU). Spoken language requires careful understanding of speaker interactions, dialog states and speech induced multimodal behaviors to generate a meaningful representation of the conversation. In this work, we propose to dissect SLU into three representative properties:conversational (disfluency, pause, overtalk), channel (speaker-type, turn-tasks) and ASR (insertion, deletion,substitution). We probe BERT based language models (BERT, RoBERTa) trained on spoken transcripts to investigate its ability to understand multifarious properties in absence of any speech cues. Empirical results indicate that LM is surprisingly good at capturing conversational properties such as pause prediction and overtalk detection from lexical tokens. On the downsides, the LM scores low on turn-tasks and ASR errors predictions. Additionally, pre-training the LM on spoken transcripts restrain its lin

Stats

• citations1
• S2 citations—
• github stars0
• HF likes0
• heat score2.26
• arxiv keykumar2021what

Related papers

• ST-BERT: Cross-modal Language Model Pre-training For End-to-end Spoken Language Understanding (2020)9.59
• Style Attuned Pre-training And Parameter Efficient Fine-tuning For Spoken Language Understanding (2020)6.77
• Multimodal Audio-textual Architecture For Robust Spoken Language Understanding (2023)0.00
• Asr-generated Text For Language Model Pre-training Applied To Speech Tasks (2022)3.58
• Wabert: A Low-resource End-to-end Model For Spoken Language Understanding And Speech-to-bert Alignment (2022)0.00
• Large Language Model Can Transcribe Speech In Multi-talker Scenarios With Versatile Instructions (2024)11.23
• Understanding Semantics From Speech Through Pre-training (2019)0.00
• Recent Advances In Speech Language Models: A Survey (2024)14.64