基于正则化的印度语ASR持续学习方法研究
A Study on Regularization-Based Continual Learning Methods for Indic ASR
AI总结：
本文针对印度语ASR，采用基于Conformer的混合RNN-T/CTC模型，在IndicSUPERB基准上评估EWC、MAS和LwF三种持续学习策略，有效缓解了多语言顺序学习中的灾难性遗忘。
摘要：
印度的语言多样性为开发包容性自动语音识别（ASR）系统带来了重大挑战。传统的多语言模型需要同时访问所有语言数据，但由于数据的顺序到达和隐私限制，这并不切实际。持续学习（CL）提供了一种解决方案，使模型能够按顺序学习新语言，而不会灾难性地遗忘先前学到的知识。本文使用 IndicSUPERB 基准的一个子集，研究了印度语言 ASR 的持续学习。我们采用基于 Conformer 的混合 RNN-T/CTC 模型，该模型最初在印地语上进行预训练，随后在另外八种印度语言上进行增量训练，形成总共九种语言的序列。我们评估了三种著名的基于正则化和蒸馏的 CL 策略：Elastic Weight Consolidation (EWC)、Memory Aware Synapses (MAS) 和 Learning without Forgetting (LwF)，选择它们是因为它们适用于无重放、注重隐私的场景。我们使用词错率（WER）分析了 RNN-T 和 CTC 路径在干净和含噪数据上的性能，以及通过向后迁移评估的知识保留情况。我们还探讨了改变每个任务训练轮数（1、2、5 和 10）的影响。与朴素微调相比，结果证明了 CL 在缓解遗忘方面的有效性，使其成为在现实约束下为多样化印度语言开发可扩展 ASR 的一种有前景的方法。代码可在以下网址获取：https://github.com/FrozenWolf-Cyber/Indic-CL-ASR
英文摘要：
Indias linguistic diversity poses significant challenges for developing inclusive Automatic Speech Recognition (ASR) systems. Traditional multilingual models, which require simultaneous access to all language data, are impractical due to the sequential arrival of data and privacy constraints. Continual Learning (CL) offers a solution by enabling models to learn new languages sequentially without catastrophically forgetting previously learned knowledge. This paper investigates CL for ASR on Indian languages using a subset of the IndicSUPERB benchmark. We employ a Conformer-based hybrid RNN-T/CTC model, initially pretrained on Hindi, which is then incrementally trained on eight additional Indian languages, for a total sequence of nine languages. We evaluate three prominent regularization- and distillation-based CL strategies: Elastic Weight Consolidation (EWC), Memory Aware Synapses (MAS), and Learning without Forgetting (LwF), selected for their suitability in no-replay, privacy-conscious scenarios. Performance is analyzed using Word Error Rate (WER) for both RNN-T and CTC paths on clean and noisy data, as well as knowledge retention via Backward Transfer. We also explore the impact of varying the number of training epochs (1, 2, 5, and 10) per task. Results, compared against naive fine-tuning, demonstrate CLs effectiveness in mitigating forgetting, making it a promising approach for scalable ASR in diverse Indian languages under realistic constraints. The code is available at: https://github.com/FrozenWolf-Cyber/Indic-CL-ASR
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。
