Hi, I’m Wenbin, a PhD Candidate at UNSW, supervised by Prof. Sanjay Jha and A/Prof. Yang Song.
My research primarily focuses on Generative AI and Text-to-Speech (TTS) technologies, with a specific interest in Voice Cloning, Talking-head Generation, and Zero-Shot Speaker-Adaptive TTS. I am also actively researching Voice Data Attribution and Anti-Spoofing within the context of trustworthy digital societies.
Beyond academia, I am the Co-founder and Technical Lead of Sozio.AI Pty Ltd, a UNSW-incubated startup where we are developing a “Virtual Tutor System” to bring real-time, interactive AI avatars to education. My work bridges the gap between cutting-edge research and commercial application, translating theoretical advancements into scalable products.
Previously, I was a Research Intern at Dolby Laboratories, working on Cross-language Speaker-Adaptive TTS and Contrastive Language-Voice Pretraining. I am currently leading the technical development for the AEA Ignite Grant project and contributing to the UNSW–UTS Trustworthy Digital Society (TDS) initiative.
Some of my research topics are outlined below. If you are interested in any of them, or potential collaborations in the generative AI space, I would be happy to discuss further. My publications and projects can also be found on my GitHub and Google Scholar pages.
Interests