Human-Centered Artificial Intelligence
I am a Research Fellow and Postdoctoral Associate at the Massachusetts Institute of Technology (January 2025–present) and a recipient of the MIT–Novo Nordisk Artificial Intelligence Postdoctoral Fellowship (2025–2027). I am mentored by Prof. Rosalind Picard and work closely with Prof. Paul Pu Liang at the MIT Media Lab.
I received my B.E. (Hons) in Telecommunications Engineering and Ph.D. in Electrical Engineering from the University of New South Wales, Sydney, Australia, in 2020 and 2024, respectively. My Ph.D. was supervised by Associate Professor Vidhyasaharan Sethu, Professor Eliathamby Ambikairajah, and Dr. Ting Dang. My research advances signal processing, machine learning, and artificial intelligence, with a particular focus on speech and multisensory intelligence for human-centered engineering.
I am the lead author of the Best Paper Award at ACII 2023, and my work was recognized as a Top 3% Paper at IEEE ICASSP 2023. I was also named a Rising Star—Women in Engineering (2023). I serve as the Publications Chair for INTERSPEECH 2026.
Research
My research advances signal processing, machine learning, and artificial intelligence through three interconnected directions:
Investigating how uncertainty, ambiguity, and subjectivity affect AI-driven decision-making in mental health, speech technology, and affective computing to improve interpretability and trustworthiness.
Developing machine learning models to track temporal changes in emotional states and mental health conditions, enabling personalized monitoring and early forecasting of emotion fluctuations, depression severity, and well-being trajectories.
Designing holistic frameworks that integrate speech with other modalities—video, text, and physiological signals—to enable more reliable and context-aware inference of human affect and behaviours that generalizes across real-world conditions.
Path
From Sydney to Cambridge, working across speech, emotion, and responsible artificial intelligence.
Postdoctoral Associate · Affective Computing
MIT–Novo Nordisk AI Postdoctoral Fellow · Host: Prof. Rosalind W. Picard
Ph.D. in Electrical Engineering · Speech Processing
Thesis: Speech Based Continuous Emotion Recognition: Modelling of Ambiguity and Temporal Dynamics ↗Check out my Three Minute Thesis competition video ↗—Runner-up at UNSW Engineering and Winner from the School of EET.B.Eng. (Hons) in Telecommunications Engineering
First Class Honours · Dean’s HonoursUpdates
Invited talk at the Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California.
Invited talk at the Agentic AI Summit 2026, UC Berkeley.
Revisiting Delay Compensation via Feature-Level Temporal Accumulation in Continuous Emotion Recognition accepted to INTERSPEECH 2026.
Hybrid Continual Learning for Low-Resource Australian Aboriginal Language Identification accepted to INTERSPEECH 2026.
Omnisapiens: A Foundation Model for Social Behavior Processing via Heterogeneity-Aware Relative Policy Optimization accepted to ICML 2026.
Human Behavior Atlas: Benchmarking Unified Psychological and Social Behavior Understanding accepted to ICLR 2026.
AMBER2: Dual Ambiguity-Aware Emotion Recognition Applied to Speech and Text accepted to ICASSP 2026.
Guest lecture for MIT MAS.S63: Affective Computing and Multimodal Interaction.
Invited talks at the University of Melbourne and UNSW on ambiguity, subjectivity, and temporal dynamics.
When One Modality Sabotages the Others: A Diagnostic Lens on Multimodal Reasoning accepted to a NeurIPS 2025 workshop.
Beyond Accuracy: A Diagnostic Protocol for Fairly Evaluating Multimodal Reasoning accepted to a NeurIPS 2025 workshop.
How Many Raters Do We Need? Analyses of Uncertainty in Estimating Ambiguity-Aware Emotion Labels accepted by IEEE Transactions on Affective Computing.
Invited talk at Novo Nordisk in Copenhagen on responsible affective AI.
Invited presentation at the IEEE International Conference on AI and Data Analytics.
Recognition
Temporal-Aware Time Series Analysis for Human Mental Behavior Modelling at the MIT Media Lab
One of 10 fellows in the cohort
One of two nominees selected by UNSW
UNSW Ph.D. scholarship
Top paper at the flagship affective computing conference
24 awardees at the Asian Deans’ Forum
Flagship signal processing conference recognition
UNSW
UNSW
Current work
IEEE Transactions on Affective Computing · 2025
IEEE Transactions on Affective Computing · 2022
Frontiers in Computer Science · 2021
arXiv · 2026
arXiv · 2026
ICLR 2026
ICASSP 2026
NeurIPS 2025 Workshop
ACII 2025
INTERSPEECH 2025
INTERSPEECH 2024
INTERSPEECH 2024
ACII 2023
🏆 Best Paper Award — ACII 2023
ICASSP 2023
⭐ Top 3% Paper Recognition — ICASSP 2023
INTERSPEECH 2023
ICASSP 2022
Community
Publication Chair
INTERSPEECH 2026 ↗Organizing Committee
Harvard-MIT Speech and Language Biomarker Network ↗Board Member
ISCA Postdoc & Early Career Researcher Advisory Committee (PECRAC) · 2026 ↗Workshop Organizer
INTERSPEECH 2026 Workshop: Challenges in Speech Data Collection, Curation, and Annotation ↗Workshop Organizer
ICASSP 2026 Workshop: Automatic Song Aesthetics Evaluation Challenge ↗Senior Panelist
INTERSPEECH Young Female Researchers in Speech Workshop · 2025 ↗Program Committee
ACII 2025 Multilingual and Multimodal Affective Computing Workshop ↗Journals: IEEE Transactions on Affective Computing; IEEE Transactions on Audio, Speech and Language Processing; Computer Speech & Language.
Conferences: EMNLP, ICASSP, INTERSPEECH, ACII, and ACIS.
Beyond research
I am a street dancer specializing in hip-hop, with extensive experience in crew performance, freestyle, and choreography across multiple styles.
I have played piano continuously since age four (China Conservatory of Music Piano Grade 9 / 中国音乐学院钢琴九级) and drum kit since age six.