aiqbench analysis
This research addresses a critical generalization gap in voice biometrics by proposing a teacher-student framework that disentangles speaker identity from spoofing cues. By combining gradient reversal layers with a Variational Information Bottleneck, the method achieves a 25.7% relative reduction in Equal Error Rate across nine datasets without requiring speaker labels. The approach offers a robust solution for deploying spoofing detection systems in out-of-domain scenarios where speaker bias typically degrades performance.
Speaker-Invariant Representation Learning for Spoofing Detection via Gradient Reversal and A Variational Information Bottleneck
arXiv:2606.08678v1 Announce Type: cross Abstract: Sophisticated generative speech technology can undermined the reliability of voice biometrics. While spoofing detection systems excel when assessed under in-domain conditions, generalisation to out-of-domain settings is often poor. In this paper, we show that such issues could be caused by speaker bias, where models learn individual voice traits rather than markers of manipulation or generation. We propose a teacher-student framework for speaker-invariant spoofing detection that disentangles identity without requiring speaker labels. We leverage a pre-trained speaker recognition teacher to guide a student model via a gradient reversal layer. To control the balance between suppressing cues related to voice identity with the preservation of those related to spoofing detection, we integrate a Variational Information Bottleneck. Evaluations across nine datasets show our model achieves a 25.7% relative reduction to the EER compared to the MHFA baseline.
Abstract:Sophisticated generative speech technology can undermined the reliability of voice biometrics. While spoofing detection systems excel when assessed under in-domain conditions, generalisation to out-of-domain settings is often poor. In this paper, we show that such issues could be caused by speaker bias, where models learn individual voice traits rather than markers of manipulation or generation. We propose a teacher-student framework for speaker-invariant spoofing detection that disentangles identity without requiring speaker labels. We leverage a pre-trained speaker recognition teacher to guide a student model via a gradient reversal layer. To control the balance between suppressing cues related to voice identity with the preservation of those related to spoofing detection, we integrate a Variational Information Bottleneck. Evaluations across nine datasets show our model achieves a 25.7% relative reduction to the EER compared to the MHFA baseline.
Sound (cs.SD); Machine Learning (cs.LG)
arXiv:2606.08678 [cs.SD]
(or arXiv:2606.08678v1 [cs.SD] for this version)
https://doi.org/10.48550/arXiv.2606.08678 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 230 words
Related coverage
Multimodal Domain Generalization for Depression Detection: An Attention-Based BiLSTM Network with Domain-Adversarial Training
arXiv · Jul 28, 2026
Data Diversity, Not Frequency Invariance: A Controlled and Self-Audited Study of Compression-Robust Deepfake Detection
arXiv · Sep 1, 2026
Links #5: 2026/07
LessWrong · Jul 31, 2026
Self-Supervised Speech Representations for Cross-Speaker Dysarthria Detection During Awake Craniotomy
arXiv · Oct 9, 2026