SAFE: Spiking Neural Network-based Audio Fidelity Evaluation
Keywords: Fake Audio Detection, Spiking Neural Networks (SNNs), Partial Fake Audio
TL;DR: Exploring the use of Spiking Neural Networks (SNNs) for detecting fake and partial fake audio.
Abstract: Recent advancements in generative AI have enabled the creation of highly realistic synthetic audio, posing significant challenges in voice authentication, media verification, and fraud detection. While deep learning models are frequently used for fake audio detection, they often struggle to generalize to unseen and complex manipulations, particularly partial fake audio, where real and synthetic segments are seamlessly combined. This paper explores the use of Spiking Neural Networks (SNNs) for fake and partial fake audio detection, an area that has not yet been investigated. SNNs, known for their energy-efficient computation and ability to process temporal data, offer a promising alternative to traditional Artificial Neural Networks (ANNs). We propose an SNN-based approach for fake audio detection and comprehensively evaluate its performance through a series of experiments, including hyperparameter tuning, cross-dataset generalization and partial fake audio detection.
Our results show that SNNs achieve accuracy comparable to state-of-the-art ANN models with fewer number of parameters. Although, SNNs did not offer significant improvements in generalization capabilities, they provided advantages such as reduced model sizes and computational efficiency, making them more suitable for resource-constrained and real-time voice authentication applications.
This study lays the groundwork for further exploration of SNNs in audio spoofing countermeasures, providing a foundation for future advancements in security-critical voice applications.
Primary Area: unsupervised, self-supervised, semi-supervised, and supervised representation learning
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 13128
Loading