Detecting Audio Deepfakes on the Edge: Lightweight SSL-Based Detection in a Browser Plugin
Abstract
Audio deepfakes are a growing challenge for the general public, as well as for journalists and fact-checkers. The latter need reliable tools to verify the authenticity of their sources, while at the same time keeping their information private. Commercial deepfake detection solutions rely on cloud-based processing, which raises privacy concerns. To solve this problem, we propose an on-device audio deepfake detection model. We show that a truncated self-supervised backbone with a simple logistic classifier is both very fast and often more accurate than existing solutions. Our solution outperforms the baseline AASIST by 10% and improves inference speed by 40%. We integrate this model into a browser plug-in, which allows journalists and fact-checkers to detect deepfakes easily and securely. Code for the plugin is available at https://github.com/OctavianPascu97/ Audio-Deepfakes-Browser-Plugin.
Citation
O. Pascu, D. Oneata, H. Cucu and N. M. Müller, "Detecting Audio Deepfakes on the Edge: Lightweight SSL-Based Detection in a Browser Plugin," 2025 International Conference on Speech Technology and Human-Computer Dialogue (SpeD), Cluj-Napoca, Romania, 2025, pp. 17-22, doi: 10.1109/SpeD67700.2025.11252312.
Read the full text
Full-text access is limited to 2 theses per day. Enter your MMCM email to open this study.
Your MMCM email is recorded only for the library’s controlled-access and borrowing records, in line with the Intellectual Property Code. It is not used for anything else.
