Manipulated Regions Localization For Partially Deepfake Audio: A Survey

TL;DR

This survey reviews the current state and challenges of localizing manipulated regions in partially deepfake audio, highlighting the need for systematic approaches to detect cryptic audio manipulations.

Contribution

It provides the first comprehensive overview of methods, limitations, and future trends in localized detection of partially deepfake audio regions.

Findings

Current methods lack robustness against sophisticated manipulations

Most approaches focus on detection rather than localization

Emerging trends aim to improve localization accuracy and reliability

Abstract

With the development of audio deepfake techniques, attacks with partially deepfake audio are beginning to rise. Compared to fully deepfake, it is much harder to be identified by the detector due to the partially cryptic manipulation, resulting in higher security risks. Although some studies have been launched, there is no comprehensive review to systematically introduce the current situations and development trends for addressing this issue. Thus, in this survey, we are the first to outline a systematic introduction for partially deepfake audio manipulated region localization tasks, including the fundamentals, branches of existing methods, current limitations and potential trends, providing a revealing insight into this scope.