Pith. sign in

REVIEW 1 cited by

Less Forgetting for Better Generalization: Exploring Continual-learning Fine-tuning Methods for Speech Self-supervised Representations

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2407.00756 v1 pith:GV5XA5VE submitted 2024-06-30 eess.AS cs.SD

classification eess.AScs.SD
keywords fine-tuningspeechduringforgettinggeneralizationself-supervisedbettercontinual-learning
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Despite being trained on massive and diverse datasets, speech self-supervised encoders are generally used for downstream purposes as mere frozen feature extractors or model initializers before fine-tuning. The former severely limits the exploitation of large encoders, while the latter hurts the robustness acquired during pretraining, especially in low-resource scenarios. This work explores middle-ground solutions, conjecturing that reducing the forgetting of the self-supervised task during the downstream fine-tuning leads to better generalization. To prove this, focusing on speech recognition, we benchmark different continual-learning approaches during fine-tuning and show that they improve both in-domain and out-of-domain generalization abilities. Relative performance gains reach 15.7% and 22.5% with XLSR used as the encoder on two English and Danish speech recognition tasks. Further probing experiments show that these gains are indeed linked to less forgetting.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. VARAN: Variational Inference for Self-Supervised Speech Models Fine-Tuning on Downstream Tasks

    cs.LG 2025-08 unverdicted novelty 4.0 of 10

    VARAN dynamically reweights internal layer representations per input through probing heads and data-dependent weights, reporting improved ASR and emotion recognition with LoRA.

Pith tools