Saved in:
Bibliographic Details
Main Authors: Yousif, Mohammed, Mathew, Jonat John, Pallan, Huzaifa, Padda, Agamjeet Singh, Shah, Syed Daniyal, Adamski, Sara, Reddiboina, Madhu, Pankajakshan, Arjun
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2404.13008
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910415261270016
author Yousif, Mohammed
Mathew, Jonat John
Pallan, Huzaifa
Padda, Agamjeet Singh
Shah, Syed Daniyal
Adamski, Sara
Reddiboina, Madhu
Pankajakshan, Arjun
author_facet Yousif, Mohammed
Mathew, Jonat John
Pallan, Huzaifa
Padda, Agamjeet Singh
Shah, Syed Daniyal
Adamski, Sara
Reddiboina, Madhu
Pankajakshan, Arjun
contents Generalization in audio deepfake detection presents a significant challenge, with models trained on specific datasets often struggling to detect deepfakes generated under varying conditions and unknown algorithms. While collectively training a model using diverse datasets can enhance its generalization ability, it comes with high computational costs. To address this, we propose a neural collapse-based sampling approach applied to pre-trained models trained on distinct datasets to create a new training database. Using ASVspoof 2019 dataset as a proof-of-concept, we implement pre-trained models with Resnet and ConvNext architectures. Our approach demonstrates comparable generalization on unseen data while being computationally efficient, requiring less training data. Evaluation is conducted using the In-the-wild dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2404_13008
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Enhancing Generalization in Audio Deepfake Detection: A Neural Collapse based Sampling and Training Approach
Yousif, Mohammed
Mathew, Jonat John
Pallan, Huzaifa
Padda, Agamjeet Singh
Shah, Syed Daniyal
Adamski, Sara
Reddiboina, Madhu
Pankajakshan, Arjun
Sound
Audio and Speech Processing
Generalization in audio deepfake detection presents a significant challenge, with models trained on specific datasets often struggling to detect deepfakes generated under varying conditions and unknown algorithms. While collectively training a model using diverse datasets can enhance its generalization ability, it comes with high computational costs. To address this, we propose a neural collapse-based sampling approach applied to pre-trained models trained on distinct datasets to create a new training database. Using ASVspoof 2019 dataset as a proof-of-concept, we implement pre-trained models with Resnet and ConvNext architectures. Our approach demonstrates comparable generalization on unseen data while being computationally efficient, requiring less training data. Evaluation is conducted using the In-the-wild dataset.
title Enhancing Generalization in Audio Deepfake Detection: A Neural Collapse based Sampling and Training Approach
topic Sound
Audio and Speech Processing
url https://arxiv.org/abs/2404.13008