Files and links (1)
Published, Version of Record (VoR)Open Access via Drexel Libraries Read and Publish Program 2025CC BY V4.0, Open
Conference proceeding
SAFE: Synthetic Audio Forensics Evaluation Challenge
IH&MMSEC '25: Proceedings of the 2025 ACM Workshop on Information Hiding and Multimedia Security, pp 174-180
17 Jun 2025
Featured in Collection : Research Supported by Drexel Libraries' OA Programs
Abstract
The increasing realism of synthetic speech generated by advanced text-to-speech (TTS) models, coupled with post-processing and laundering techniques, presents a significant challenge for audio forensic detection. In this paper, we introduce the SAFE (Synthetic Audio Forensics Evaluation) Challenge, a fully blind evaluation framework designed to benchmark detection models across progressively harder scenarios: raw synthetic speech, processed audio (e.g., compression, resampling), and laundered audio intended to evade forensic analysis. The SAFE challenge consisted of a total of 90 hours of audio and 21k audio samples split across 21 different real sources and 17 different TTS models and 3 tasks. We present the challenge, evaluation design and tasks, dataset details, and initial insights into the strengths and limitations of current approaches, offering a foundation for advancing synthetic audio detection research. More information is available at https://stresearch.github.io/SAFE/.
Metrics
7 Record Views
Details
- Title
- SAFE: Synthetic Audio Forensics Evaluation Challenge
- Creators
- Kirill Trapeznikov - Systems and Technology Research (United States)Paul Cummer - Systems and Technology Research (United States)Pranay Pherwani - Systems and Technology Research (United States)Jai Aslam - Systems and Technology Research (United States)Michael Davinroy - Aptima (United States)Peter Bautista - Aptima (United States)Laura Cassani - Aptima (United States)Matthew C Stamm (Corresponding Author) - Drexel University, Electrical and Computer Engineering
- Publication Details
- IH&MMSEC '25: Proceedings of the 2025 ACM Workshop on Information Hiding and Multimedia Security, pp 174-180
- Conference
- 13th ACM Workshop on Information Hiding and Multimedia Security, IHandMMSec 2025 (San Jose, California, United States, 18 Jun 2025–20 Jun 2025)
- Publisher
- Association for Computing Machinery
- Number of pages
- 7
- Grant note
- This material is based on research sponsored by ULRI DSRI.
- Resource Type
- Conference proceeding
- Language
- English
- Academic Unit
- Electrical and Computer Engineering
- Web of Science ID
- WOS:001540643000019
- Scopus ID
- 2-s2.0-105031783920
- Other Identifier
- 991022063442204721
InCites Highlights
Data related to this publication, from InCites Benchmarking & Analytics tool:
- Collaboration types
- Industry collaboration
- Domestic collaboration
- Web of Science research areas
- Computer Science, Information Systems
- Computer Science, Theory & Methods
- Mathematics, Applied