Leave No Knowledge Behind During Knowledge Distillation: Towards Practical and Effective Knowledge Distillation for Code-Switching ASR Using Realistic Data

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tseng, Liang-Hsuan, Chen, Zih-Ching, Chang, Wei-Shun, Lee, Cheng-Kuang, Huang, Tsung-Ren, Lee, Hung-yi
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909255328595968
author Tseng, Liang-Hsuan
Chen, Zih-Ching
Chang, Wei-Shun
Lee, Cheng-Kuang
Huang, Tsung-Ren
Lee, Hung-yi
author_facet Tseng, Liang-Hsuan
Chen, Zih-Ching
Chang, Wei-Shun
Lee, Cheng-Kuang
Huang, Tsung-Ren
Lee, Hung-yi
contents Recent advances in automatic speech recognition (ASR) often rely on large speech foundation models for generating high-quality transcriptions. However, these models can be impractical due to limited computing resources. The situation is even more severe in terms of more realistic or difficult scenarios, such as code-switching ASR (CS-ASR). To address this, we present a framework for developing more efficient models for CS-ASR through knowledge distillation using realistic speech-only data. Our proposed method, Leave No Knowledge Behind During Knowledge Distillation (K$^2$D), leverages both the teacher model's knowledge and additional insights from a small auxiliary model. We evaluate our approach on two in-domain and two out-domain datasets, demonstrating that K$^2$D is effective. By conducting K$^2$D on the unlabeled realistic data, we have successfully obtained a 2-time smaller model with 5-time faster generation speed while outperforming the baseline methods and the teacher model on all the testing sets. We have made our model publicly available on Hugging Face (https://huggingface.co/andybi7676/k2d-whisper.zh-en).
format Preprint
id arxiv_https___arxiv_org_abs_2407_10603
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Leave No Knowledge Behind During Knowledge Distillation: Towards Practical and Effective Knowledge Distillation for Code-Switching ASR Using Realistic Data
Tseng, Liang-Hsuan
Chen, Zih-Ching
Chang, Wei-Shun
Lee, Cheng-Kuang
Huang, Tsung-Ren
Lee, Hung-yi
Audio and Speech Processing
Computation and Language
Sound
Recent advances in automatic speech recognition (ASR) often rely on large speech foundation models for generating high-quality transcriptions. However, these models can be impractical due to limited computing resources. The situation is even more severe in terms of more realistic or difficult scenarios, such as code-switching ASR (CS-ASR). To address this, we present a framework for developing more efficient models for CS-ASR through knowledge distillation using realistic speech-only data. Our proposed method, Leave No Knowledge Behind During Knowledge Distillation (K$^2$D), leverages both the teacher model's knowledge and additional insights from a small auxiliary model. We evaluate our approach on two in-domain and two out-domain datasets, demonstrating that K$^2$D is effective. By conducting K$^2$D on the unlabeled realistic data, we have successfully obtained a 2-time smaller model with 5-time faster generation speed while outperforming the baseline methods and the teacher model on all the testing sets. We have made our model publicly available on Hugging Face (https://huggingface.co/andybi7676/k2d-whisper.zh-en).
title Leave No Knowledge Behind During Knowledge Distillation: Towards Practical and Effective Knowledge Distillation for Code-Switching ASR Using Realistic Data
topic Audio and Speech Processing
Computation and Language
Sound
url https://arxiv.org/abs/2407.10603