I am currently a PhD student at the University of Melbourne in Australia. My research journey is fully supported by the Melbourne Research Scholarship, and I’m incredibly fortunate to be supervised by Dr. Ting Dang and Prof. Eun-Jung Holden. Before joining Unimelb, I was an AI engineer at Fortemedia working with Dr. Rohan Kumar Das. I graduated as a CS Master student at NTU advised by Prof. Chng Eng Siong. I obtained my B.Eng degree from Jilin University.

I am contributing to building “adaptive”, “efficient”, and “robust” next-generation speech AI systems. At this moment, I mainly work in the post training paradigms for speech learning systems. Specifically, by merging continual learning, domain adaptation, knowledge editing, and reinforcement fine-tuning, we pave the way for speech models that continuously adapt, specialize efficiently, and self-correct in real-world environments. I have published more than 20 papers at the top international AI conferences and journals such as ACL, SPL, ICME, ICASSP, and INTERSPEECH.

🔥 News

• 2025.09:  🎉🎉 Our ICASSP grand challenge ESDD 2026 has been launched.
• 2025.08:  🎉🎉 I joined the University of Melbourne as a PhD student in Australia!
• 2025.05:  🎉🎉 Four papers have been accepted to Interspeech 2025!

🔍 Research Area

Speech and Audio Processing: Sound Event Detection, Spoken Keyword Spotting, Speech Foundation Model, DeepFake Detection

Algorithm: Continual learning, Test time adaptation, Knowledge editing

🎓 Education

• 08.2025 - Now, Doctor of Philosophy - Engineering and IT, The University of Melbourne, Australia
• 08.2021 - 01.2023, Master of Science (Artificial Intelligence), Nanyang Technological University, Singapore
• 08.2016 - 07.2020, B.E. in Internet of Things Engineering, Jilin University, Changchun, China

💼 Work Experience

• 01.2023 - 08.2025, AI Engineer, Fortemedia Singapore
• 07.2020 - 05.2021, Software Engineer, China Mobile (Chengdu) Industrial Research Institute

📝 Publications

INTERSPEECH 2025

EnvSDD: Benchmarking Environmental Sound Deepfake Detection

Han Yin, Yang Xiao, Rohan Kumar Das, Jisheng Bai, Haohe Liu, Wenwu Wang, Mark D Plumbley

Project

• The first large-scale curated dataset designed for Environmental Sound Deepfake Detection.

Continual Learning for Speech / Audio

• INTERSPEECH 2025 Listen, Analyze, and Adapt to Learn New Attacks: An Exemplar-Free Class Incremental Learning Method for Audio Deepfake Source Tracing, Yang Xiao, Rohan Kumar Das.
• ACL 2025 AnalyticKWS: Towards Exemplar-Free Analytic Class Incremental Learning for Small-footprint Keyword Spotting, Yang Xiao, Peng Tianyi, Rohan Kumar Das, Yuchen Hu, Huiping Zhuang
• ICME 2025 Where’s That Voice Coming? Continual Learning for Sound Source Localization, Yang Xiao, Rohan Kumar Das.
• ICASSP 2025 UCIL: An Unsupervised Class Incremental Learning Approach for Sound Event Detection, Yang Xiao, Rohan Kumar Das.
• ICASSP 2025 Dark Experience for Incremental Keyword Spotting, Tianyi Peng, Yang Xiao.
• DCASE 2022 Continual Learning For On-Device Environmental Sound Classification, Yang Xiao*, Xubo Liu*, James King, Arshdeep Singh, Eng Siong Chng, Mark D. Plumbley, Wenwu Wang.
• INTERSPEECH 2022 Rainbow Keywords: Efficient Incremental Learning for Online Spoken Keyword Spotting, Yang Xiao, Nana Hou, Eng Siong Chng.

Domain adaptation for Speech / Audio

• INTERSPEECH 2025 AdaKWS: Towards Robust Keyword Spotting with Test-Time Adaptation, Yang Xiao, Tianyi Peng, Yanghao Zhou, Rohan Kumar Das.
• APSIPA ASC 2025 DG-SED: Domain Generalization for Sound Event Detection with Heterogeneous Training Data, Yang Xiao, Han Yin, Jisheng Bai, Rohan Kumar Das.
• DCASE 2024 WildDESED: An LLM-Powered Dataset for Wild Domestic Environment Sound Event Detection System, Yang Xiao, Rohan Kumar Das.

Others

2025

• INTERSPEECH 2025 TF-Mamba: A Time-Frequency Network for Sound Source Localization, Yang Xiao, Rohan Kumar Das.
• SPSC 2025Multilingual Source Tracing of Speech Deepfakes: A First Benchmark, Xi Xuan, Yang Xiao, Rohan Kumar Das, Tomi Kinnunen.
• SPL 2025 XLSR-Mamba: A Dual-Column Bidirectional State Space Model for Spoofing Attack Detection, Yang Xiao, Rohan Kumar Das.
• APSIPA ASC 2025 RawTFNet: A Lightweight CNN Architecture for Speech Anti-spoofing, Yang Xiao, Ting Dang, Rohan Kumar Das.

Before 2024

• INTERSPEECH 2023 Small Footprint Multi-channel Network for Keyword Spotting with Centroid Based Awareness, Dianwen Ng, Yang Xiao, Jia Qi Yip, Zhao Yang, Biao Tian, Qiang Fu, Eng Siong Chng, Bin Ma.

🎖 Honors and Awards

• 2025.07 ISCA (International Speech Communication Association) Grant, Interspeech, Rotterdam
• 2025.03 Melbourne Research Scholarship, University of Melbourne