Is Smaller Always Faster? Tradeoffs in Compressing Self-Supervised Speech Transformers
Journal
ASRU 2025 - 2025 IEEE Automatic Speech Recognition and Understanding Workshop
Start Page
1
End Page
7
ISBN (of the container)
979-833154426-3
ISBN
[9798331544263]
Date Issued
2025-12-06
Author(s)
Abstract
Transformer-based self-supervised models have achieved remarkable success in speech processing, but their large size and high inference cost present significant challenges for real-world deployment. While numerous compression techniques have been proposed, inconsistent evaluation metrics make it difficult to compare their practical effectiveness. In this work, we conduct a comprehensive study of four common compression methods, including weight pruning, head pruning, low-rank approximation, and knowledge distillation on self-supervised speech Transformers. We evaluate each method under three key metrics: parameter count, multiply-accumulate operations, and real-time factor. Results show that each method offers distinct advantages. In addition, we contextualize recent compression techniques, comparing DistilHuBERT, FitHuBERT, LightHuBERT, ARMHuBERT, and STaRHuBERT under the same framework, offering practical guidance on compression for deployment.
Event(s)
2025 IEEE Automatic Speech Recognition and Understanding Workshop, ASRU 2025
Subjects
model compression
Speech self-supervised model
Publisher
Institute of Electrical and Electronics Engineers(IEEE)
Type
conference paper