Improving Multilingual Speech Models on ML-SUPERB 2.0: Fine-tuning with Data Augmentation and LID-Aware CTC

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Wang, Qingzheng, Sun, Jiancheng, Peng, Yifan, Watanabe, Shinji
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916774062063616
author Wang, Qingzheng
Sun, Jiancheng
Peng, Yifan
Watanabe, Shinji
author_facet Wang, Qingzheng
Sun, Jiancheng
Peng, Yifan
Watanabe, Shinji
contents Multilingual speech processing with self-supervised or supervised pre-trained Speech Foundation Models (SFM) has achieved strong performance on tasks like Language Identification (LID) and Automatic Speech Recognition (ASR). However, these models struggle with limited resources during fine-tuning. This paper enhances multilingual LID and ASR on ML-SUPERB 2.0 by exploring multiple strategies for adapting SFMs, including frozen upstream training, partial fine-tuning, and low-rank adaptation. Furthermore, we employ data augmentation to mitigate performance gaps in few-shot settings and introduce LID Connectionist Temporal Classification (CTC) loss for regularization. Our approach achieves a 14% relative improvement in LID accuracy and a 30% relative reduction in ASR CER over the baseline on ML-SUPERB 2.0, securing second place in the Interspeech 2025 ML-SUPERB 2.0 Challenge.
format Preprint
id arxiv_https___arxiv_org_abs_2505_24200
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Improving Multilingual Speech Models on ML-SUPERB 2.0: Fine-tuning with Data Augmentation and LID-Aware CTC
Wang, Qingzheng
Sun, Jiancheng
Peng, Yifan
Watanabe, Shinji
Sound
Computation and Language
Audio and Speech Processing
Multilingual speech processing with self-supervised or supervised pre-trained Speech Foundation Models (SFM) has achieved strong performance on tasks like Language Identification (LID) and Automatic Speech Recognition (ASR). However, these models struggle with limited resources during fine-tuning. This paper enhances multilingual LID and ASR on ML-SUPERB 2.0 by exploring multiple strategies for adapting SFMs, including frozen upstream training, partial fine-tuning, and low-rank adaptation. Furthermore, we employ data augmentation to mitigate performance gaps in few-shot settings and introduce LID Connectionist Temporal Classification (CTC) loss for regularization. Our approach achieves a 14% relative improvement in LID accuracy and a 30% relative reduction in ASR CER over the baseline on ML-SUPERB 2.0, securing second place in the Interspeech 2025 ML-SUPERB 2.0 Challenge.
title Improving Multilingual Speech Models on ML-SUPERB 2.0: Fine-tuning with Data Augmentation and LID-Aware CTC
topic Sound
Computation and Language
Audio and Speech Processing
url https://arxiv.org/abs/2505.24200