Research Group

Efficient machine learning, human-centered AI, and mobile intelligence.

Research Focus

We work at the intersection of efficient ML, human-centered AI, and mobile intelligence—bringing capable AI to devices and wearables while preserving privacy.

  • Efficient & on-device ML: Small language and vision-language models, test-time adaptation, efficient inference.
  • Human-centered sensing: Personalized sensing, mobile and wearable health, affective computing.
  • Mobile intelligence: On-device agents, embodied AI, foundation models for ubiquitous computing.

Highlights

  • Publications at top venues including NeurIPS, ACM UbiComp, ACM MobiCom, ACM IPSN, IEEE CIKM, and ACM SenSys.
  • Recognition including ACM MobiSys Rising Star Award (SIGMOBILE), Best Paper Runner-up @ ACM IPSN, Best Poster @ ACM MobiCom, and multiple Best Presentation / High Distinction Thesis awards.
  • Collaborations with leading universities and industry; alumni in academia at prestigious institutions and top PhD programs.

PhD Students

Le Fang
University of Melbourne (2024-)
Co-supervising with Prof. Jorge Goncalves and Prof. Vassilis Kostakos
Mobile Phone Over Usage Detection via On-Device LLMs
ACM UbiComp 24'
Yuanbo Tong
University of Auckland (Incoming)
Co-supervising with Dr. Xinyu Zhang and A/Prof. Jing Sun
Qicong Hu
University of Auckland & Southwest University
Visiting PhD student
Xin Wang
University of Auckland (Incoming)
Co-supervising with Dr. Xinyu Zhang
Zhengying Liang
University of Auckland & Beijing Jiaotong University
Visiting PhD student

MPhil Students

  • Letian Chen (University of Melbourne (2025-), primary supervising) — Efficient Inference Engine of LLMs
  • Chathuri Seneviratne (University of Auckland (2025-), primary supervising) — Applied Medical Agent
  • Puyue Wang (University of Auckland (2025-), primary supervising) — Efficient VLA and Embodied Agents
  • Hang Song (University of Melbourne (2025-), co-supervising with Dr. Lingyan Ruan) — Efficient and Robust Guassian Splatting for 3D Scene Reconstruction
  • Songqi Guo (University of Auckland (2025-), primary supervising) — Efficient Multi-Modal LLMs for Emotion Monitoring
  • Qiuchi (Joyce) Hu (University of Melbourne (2025-), co-supervising) — Knowledge-editing for Audio LLMs
  • Sung Kyun Chung (University of Melbourne (2025-), co-supervising) — Knowledge-editing for LLMs
  • Hanwen Jiang (University of Melbourne (2024-), principal supervising) — Efficient and Robust Tiny VLMs

Mentoring

Jiaheng Dong
University of Melbourne (2025-)
Mentoring with Dr. Ting Dang
Test-time Adaptation for Speech Recognition
NeurIPS 25'

Alumni

  • Shiquan Zhang (University of Melbourne (2024)) ACM UbiComp 24'
  • Yu (Yvonne) Wu (University of Cambridge (2022-2024)) → PostDoc at Dartmouth College DeepMind Travel Award @ MLHC 24' IEEE CIKM 24' MLHC 24' NeurIPS Adjunct 23'
  • Yanxiang Wang (UNSW (2021-2024)) → Assistant Professor at Shandong University ACM UbiComp 23' ACM IPSN 23' ACM MobiCom 22'
  • Cynthia Dong (Mphil, University of Cambridge, co-supervised with Prof. Cecilia Mascolo) → PhD Award at University of Washington High Distinction Thesis @ University of Cambridge
  • Tianyi (Tina) Zhang (University of Melbourne (2024-)) → University of Sydney Best Paper Presentation Award @ ACM EIFCom 24' Best Poster Award @ CIS Doctoral Colloquium 24'
    ACM UbiComp Adjunct 24' ACM MobiCom 24' Adjunct
  • Jiawei Hu (UNSW (2021-2024)) — Toward AI-Powered Adaptive Visible Light Positioning Best Paper Runner-up Award @ ACM IPSN 24' Best Poster Award @ ACM MobiCom 22' Best PhD Forum Presentation Award @ ACM SenSys 24'
    ACM UbiComp 23' ACM IPSN 23' ACM MobiCom 22'
  • Mark Cardamis (UNSW (2023-2024)) — Smart Agriculture Using mmWave Sensing and Artificial Intelligence IEEE Internet of Things Journal 24'
  • Andy Wang (MSc, University of Melbourne, principal supervising) — On-Device Test-time Adaptation Using VLMs High Distinction Thesis @ University of Melbourne
  • Xin Wang (University of Melbourne (2024-), principal supervising) — Personalized Healthcare Monitoring via Mobile LLMs ACM MobiCom 24' Adjunct NeurIPS GenAI4Health 25' High Distinction Thesis @ University of Melbourne
  • ACM UbiComp 23' High Distinction Thesis @ Shenzhen University
  • Zhuoran Wei (MSc, University of Melbourne, co-supervising with Dr. Andrew Irlitti) — Gesture-based locomotion techniques in virtual reality High Distinction Thesis @ University of Melbourne

Collaborations

University of Auckland University of Auckland University of Melbourne University of Melbourne University of Cambridge University of Cambridge UNSW Sydney UNSW Sydney University of Oxford University of Oxford University of Adelaide University of Adelaide Dartmouth College Dartmouth College University of Sydney University of Sydney University of Birmingham University of Birmingham Macquarie University Macquarie University CSIRO CSIRO Google Google Samsung AI Samsung AI Nokia Bell Labs Nokia Bell Labs Flower Labs Flower Labs