SPSC symposium
News
Four PSST! Fellows presented their work at the 6th Symposium on Security and Privacy in Speech Communication (SPSC), held at UNSW Sydney, Australia, on September 26, 2026.
The symposium brings together researchers and practitioners from speech technology, security, law and ethics to discuss how to protect privacy and strengthen security in speech communication. This year’s event was held jointly with the VoicePrivacy Challenge 2026 workshop.
The Fellows’ contributions included:
- Chenyi Lin presented Non-Uniform Privacy Leakage: Subgroup Disparities in Membership Inference Attacks on Speaker Recognition Systems.
- Dāvis Šterns presented Goodbye Equal Error Rate, Hello Local Information Disclosure: Evaluating Voice Anonymisation against 1-to-N Linkage Threats.
- Théo Nguyen presented a poster, PSST1 System Description for VoicePrivacy Challenge 2026, co-authored with fellow Ben Luks, Tom Bäckström and Rainer Martin.
- Jiusi Zheng joined remotely to present Towards Interpretable Speaker Representations for Controllable Voice Privacy Protection and Radboud-Aalto Submission to the Voice Privacy Challenge 2026.
