Papers
arxiv:2506.16574

Weight Factorization and Centralization for Continual Learning in Speech Recognition

Published on Jun 19
Authors:
,
,

Abstract

A continual learning approach with factorization and centralization phases prevents catastrophic forgetting in multilingual speech recognition models by accumulating knowledge in low-rank adapters.

AI-generated summary

Modern neural network based speech recognition models are required to continually absorb new data without re-training the whole system, especially in downstream applications using foundation models, having no access to the original training data. Continually training the models in a rehearsal-free, multilingual, and language agnostic condition, likely leads to catastrophic forgetting, when a seemingly insignificant disruption to the weights can destructively harm the quality of the models. Inspired by the ability of human brains to learn and consolidate knowledge through the waking-sleeping cycle, we propose a continual learning approach with two distinct phases: factorization and centralization, learning and merging knowledge accordingly. Our experiments on a sequence of varied code-switching datasets showed that the centralization stage can effectively prevent catastrophic forgetting by accumulating the knowledge in multiple scattering low-rank adapters.

Community

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2506.16574 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2506.16574 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2506.16574 in a Space README.md to link it from this page.

Collections including this paper 1