← Search

Zhaolin Wei

2 accepted papers

2026

Time Shuffle: A Transferability-Booster for Multiple Audio Adversarial Tasks

AAAI 2026technical

Existing audio adversarial attack methods suffer from poor transferability, primarily due to insufficient exploration of model decision mechanisms and overreliance on heuristic-driven algorithm design. This paper aims to alleviate this gap. Specifically, through observations across three mainstream

Cited by 0SourcePDFScholar
2025

From Voices to Beats: Enhancing Music Deepfake Detection by Identifying Forgeries in Background

ICASSP 2025accepted

Music deepfake detection is aimed at identifying whether songs are generated by AI. Current methods usually separate vocals from background music for detection, but this could leave residual forgery information in the background. Our study demonstrates for the first time that incorporating backgroun…

Cited by 0SourceScholar