Saved in:
Bibliographic Details
Main Authors: Cheng, Bo, Cao, Songjun, Zhang, Xiaoming, Chen, Jie, Ma, Long, Chen, Fei
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.26465
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909000670380032
author Cheng, Bo
Cao, Songjun
Zhang, Xiaoming
Chen, Jie
Ma, Long
Chen, Fei
author_facet Cheng, Bo
Cao, Songjun
Zhang, Xiaoming
Chen, Jie
Ma, Long
Chen, Fei
contents Achieving robust generalization against unseen attacks remains a challenge in Audio Deepfake Detection (ADD), driven by the rapid evolution of generative models. To address this, we propose a framework centered on hard sample classification. The core idea is that a model capable of distinguishing challenging hard samples is inherently equipped to handle simpler cases effectively. We investigate multiple reconstruction paradigms, identifying the diffusion-based method as optimal for generating hard samples. Furthermore, we leverage multi-layer feature aggregation and introduce a Regularization-Assisted Contrastive Learning (RACL) objective to enhance generalizability. Experiments demonstrate the superior generalization of our approach, with our best model achieving a significant reduction in the average Equal Error Rate (EER) compared to the baseline.
format Preprint
id arxiv_https___arxiv_org_abs_2604_26465
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Diffusion Reconstruction towards Generalizable Audio Deepfake Detection
Cheng, Bo
Cao, Songjun
Zhang, Xiaoming
Chen, Jie
Ma, Long
Chen, Fei
Sound
Achieving robust generalization against unseen attacks remains a challenge in Audio Deepfake Detection (ADD), driven by the rapid evolution of generative models. To address this, we propose a framework centered on hard sample classification. The core idea is that a model capable of distinguishing challenging hard samples is inherently equipped to handle simpler cases effectively. We investigate multiple reconstruction paradigms, identifying the diffusion-based method as optimal for generating hard samples. Furthermore, we leverage multi-layer feature aggregation and introduce a Regularization-Assisted Contrastive Learning (RACL) objective to enhance generalizability. Experiments demonstrate the superior generalization of our approach, with our best model achieving a significant reduction in the average Equal Error Rate (EER) compared to the baseline.
title Diffusion Reconstruction towards Generalizable Audio Deepfake Detection
topic Sound
url https://arxiv.org/abs/2604.26465