arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2606.09342eess.AS
收藏

参数高效持续学习用于自动语音识别

Parameter-Efficient Continual Learning for Automatic Speech Recognition

KU Leuven(荷语鲁汶大学)

Steven Vander Eeckt, Hugo Van hamme

AI总结:

提出一种基于奇异值分解的头部-尾部子空间划分方法,在低能量尾部子空间进行近似旋转适应,结合权重平均减少遗忘,实现参数高效的持续学习ASR。

摘要:

语音基础模型能够实现强大的通用自动语音识别(ASR),并且对于下游适应具有吸引力。然而,它们的规模以及顺序微调导致的灾难性遗忘要求参数高效和正则化的训练方法,从而推动了参数高效持续学习(PECL)。虽然PECL在自然语言处理和视觉领域已被广泛研究,但在ASR中受到的关注较少。在本文中,我们基于最近在ASR参数高效微调方面的进展,提出了一种简单而有效的PECL方法。我们根据奇异值将预训练权重矩阵划分为头部和尾部子空间,并将适应限制在低能量尾部子空间内的近似旋转,从而保留主导成分并减少遗忘。对于后续任务,通过权重平均组合旋转以进一步提高保留能力。在两个基准上的实验表明,与最近的PECL基线相比,该方法减少了遗忘并实现了更优的整体性能。

英文摘要:

Speech foundation models enable strong general-purpose ASR and are attractive for downstream adaptation. However, their size and the catastrophic forgetting induced by sequential fine-tuning demand parameter-efficient and regularized training methods, motivating parameter-efficient continual learning (PECL). While PECL has been widely studied in NLP and vision, it has received less attention in ASR. In this paper, we propose a simple yet effective PECL method based on recent advances in parameter-efficient fine-tuning for ASR. We partition pretrained weight matrices into head and tail subspaces according to singular values and restrict adaptation to approximate rotations within the low-energy tail subspace, preserving dominant components and reducing forgetting. For subsequent tasks, rotations are combined via weight averaging to further improve retention. Experiments on two benchmarks demonstrate reduced forgetting and superior overall performance compared to recent PECL baselines.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑