Audio Contrastive-based Fine-tuning: Decoupling Representation Learning and Classification

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Yang, Liang, Qibin, Xiao, Chenghao, Li, Yizhi, Moubayed, Noura Al, Lin, Chenghua
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912597320663040
author Wang, Yang
Liang, Qibin
Xiao, Chenghao
Li, Yizhi
Moubayed, Noura Al
Lin, Chenghua
author_facet Wang, Yang
Liang, Qibin
Xiao, Chenghao
Li, Yizhi
Moubayed, Noura Al
Lin, Chenghua
contents Standard fine-tuning of pre-trained audio models couples representation learning with classifier training, which can obscure the true quality of the learned representations. In this work, we advocate for a disentangled two-stage framework that separates representation refinement from downstream evaluation. First, we employ a "contrastive-tuning" stage to explicitly improve the geometric structure of the model's embedding space. Subsequently, we introduce a dual-probe evaluation protocol to assess the quality of these refined representations from a geometric perspective. This protocol uses a linear probe to measure global linear separability and a k-Nearest Neighbours probe to investigate the local structure of class clusters. Our experiments on a diverse set of audio classification tasks show that our framework provides a better foundation for classification, leading to improved accuracy. Our newly proposed dual-probing framework acts as a powerful analytical lens, demonstrating why contrastive learning is more effective by revealing a superior embedding space. It significantly outperforms vanilla fine-tuning, particularly on single-label datasets with a large number of classes, and also surpasses strong baselines on multi-label tasks using a Jaccard-weighted loss. Our findings demonstrate that decoupling representation refinement from classifier training is a broadly effective strategy for unlocking the full potential of pre-trained audio models. Our code will be publicly available.
format Preprint
id arxiv_https___arxiv_org_abs_2309_11895
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Audio Contrastive-based Fine-tuning: Decoupling Representation Learning and Classification
Wang, Yang
Liang, Qibin
Xiao, Chenghao
Li, Yizhi
Moubayed, Noura Al
Lin, Chenghua
Sound
Artificial Intelligence
Computation and Language
Audio and Speech Processing
Standard fine-tuning of pre-trained audio models couples representation learning with classifier training, which can obscure the true quality of the learned representations. In this work, we advocate for a disentangled two-stage framework that separates representation refinement from downstream evaluation. First, we employ a "contrastive-tuning" stage to explicitly improve the geometric structure of the model's embedding space. Subsequently, we introduce a dual-probe evaluation protocol to assess the quality of these refined representations from a geometric perspective. This protocol uses a linear probe to measure global linear separability and a k-Nearest Neighbours probe to investigate the local structure of class clusters. Our experiments on a diverse set of audio classification tasks show that our framework provides a better foundation for classification, leading to improved accuracy. Our newly proposed dual-probing framework acts as a powerful analytical lens, demonstrating why contrastive learning is more effective by revealing a superior embedding space. It significantly outperforms vanilla fine-tuning, particularly on single-label datasets with a large number of classes, and also surpasses strong baselines on multi-label tasks using a Jaccard-weighted loss. Our findings demonstrate that decoupling representation refinement from classifier training is a broadly effective strategy for unlocking the full potential of pre-trained audio models. Our code will be publicly available.
title Audio Contrastive-based Fine-tuning: Decoupling Representation Learning and Classification
topic Sound
Artificial Intelligence
Computation and Language
Audio and Speech Processing
url https://arxiv.org/abs/2309.11895