Saved in:
Bibliographic Details
Main Authors: Chen, Liangyu, Yue, Zihao, Xu, Boshen, Jin, Qin
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2409.06709
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929495891509248
author Chen, Liangyu
Yue, Zihao
Xu, Boshen
Jin, Qin
author_facet Chen, Liangyu
Yue, Zihao
Xu, Boshen
Jin, Qin
contents Audio-Visual Source Localization (AVSL) aims to localize the source of sound within a video. In this paper, we identify a significant issue in existing benchmarks: the sounding objects are often easily recognized based solely on visual cues, which we refer to as visual bias. Such biases hinder these benchmarks from effectively evaluating AVSL models. To further validate our hypothesis regarding visual biases, we examine two representative AVSL benchmarks, VGG-SS and EpicSounding-Object, where the vision-only models outperform all audiovisual baselines. Our findings suggest that existing AVSL benchmarks need further refinement to facilitate audio-visual learning.
format Preprint
id arxiv_https___arxiv_org_abs_2409_06709
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Unveiling Visual Biases in Audio-Visual Localization Benchmarks
Chen, Liangyu
Yue, Zihao
Xu, Boshen
Jin, Qin
Multimedia
Artificial Intelligence
Sound
Audio and Speech Processing
Audio-Visual Source Localization (AVSL) aims to localize the source of sound within a video. In this paper, we identify a significant issue in existing benchmarks: the sounding objects are often easily recognized based solely on visual cues, which we refer to as visual bias. Such biases hinder these benchmarks from effectively evaluating AVSL models. To further validate our hypothesis regarding visual biases, we examine two representative AVSL benchmarks, VGG-SS and EpicSounding-Object, where the vision-only models outperform all audiovisual baselines. Our findings suggest that existing AVSL benchmarks need further refinement to facilitate audio-visual learning.
title Unveiling Visual Biases in Audio-Visual Localization Benchmarks
topic Multimedia
Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2409.06709