vix.ing · top · new · best · stats · spec

Continual Learning for End-to-End ASR by Averaging Domain Experts

2023/05/12 by Peter Plantinga, Plantinga, Peter, Jaekwon Yoo +3
Computer Science · #Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Music and Audio Processing #Natural Language Processing Techniques #Sound (cs.SD) #Speech Recognition and Synthesis #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2305.09681

openalex publication_date 2023/05/12 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Continual learning for end-to-end automatic speech recognition has to contend with a number of difficulties. Fine-tuning strategies tend to lose performance on data already seen, a process known as catastrophic forgetting. On the other hand, strategies that freeze parameters and append tunable parameters must maintain multiple models. We suggest a strategy that maintains only a single model for inference and avoids catastrophic forgetting. Our experiments show that a simple linear interpolation of several models' parameters, each fine-tuned from the same generalist model, results in a single model that performs well on all tested data. For our experiments we selected two open-source end-to-end speech recognition models pre-trained on large datasets and fine-tuned them on 3 separate datasets: SGPISpeech, CORAAL, and DiPCo. The proposed average of domain experts model performs well on all tested data, and has almost no loss in performance on data from the domain of original training.

Related