Renhao Huang

I am the Postdoc Research Fellow and the associate director for AI research at iCinema Research Centre, UNSW, focusing on developing AI models for iFire project. In this project, I mainly explore deep learning–based approaches to wildfire behaviour modelling, bridging advanced computation with urgent real-world challenges.

In 2020, I received the B.E degree with a first-class honour from the University of New South Wales. In 2025, I received my PhD degree at the School of Computer Science and Engineering, UNSW, with my thesis Towards Robust and Lightweight Vision-based Pedestrian Trajectory Prediction.

My current research focuses on:

Human Trajectory Prediction
Computer Vision
Domain Adaptation & Generalisation
Wildfire Research

News

April, 2026	One proposal paper accepted by IJCAI 2026 AI for Goods.
April, 2026	One abstract paper accepted by X-Fire 2026.
July, 2025	One abstract accepted by X-Fire, 1st International Workshop on Extreme Wildfire Events
March, 2025	PhD Thesis accepted.
March, 2025	One paper accepted by TNNLS.

Selected Publications

Vision-based Multimodal trajectory prediction: A survey

Renhao Huang, Hao Xue, Maurice Pagnucco, and 2 more authors

IEEE Transactions on Neural Networks and Learning Systems, 2025

[ Bib ] [ Paper ]

@article{survey-mtp-arxiv,
  title = {Vision-based Multimodal trajectory prediction: A survey},
  author = {Huang, Renhao and Xue, Hao and Pagnucco, Maurice and Salim, Flora and Song, Yang},
  journal = {IEEE Transactions on Neural Networks and Learning Systems},
  year = {2025},
}

Fully Decoupling Trajectory and Scene Encoding for Lightweight Heatmap-oriented Trajectory Prediction

Renhao Huang, Jingtao Ding, Maurice Pagnucco, and 1 more author

IEEE Robotics and Automation Letters, 2024

[ Bib ] [ Paper ]

@article{decouple-traj,
  author = {Huang, Renhao and Ding, Jingtao and Pagnucco, Maurice and Song, Yang},
  journal = {IEEE Robotics and Automation Letters},
  title = {{Fully Decoupling Trajectory and Scene Encoding for Lightweight Heatmap-oriented Trajectory Prediction}},
  year = {2024},
  pages = {9143--9150},
}

HyperTraj: Towards Simple and Fast Scene-compliant Endpoint Conditioned Trajectory Prediction

Renhao Huang, Maurice Pagnucco, and Yang Song

In IEEE/RSJ International Conference on Intelligent Robots and Systems, 2023

[ Bib ] [ Paper ]

@inproceedings{hypertraj,
  title = {{HyperTraj: Towards Simple and Fast Scene-compliant Endpoint Conditioned Trajectory Prediction}},
  author = {Huang, Renhao and Pagnucco, Maurice and Song, Yang},
  booktitle = {IEEE/RSJ International Conference on Intelligent Robots and Systems},
  year = {2023},
}

Towards Single Source Domain Generalisation in Trajectory Prediction: A Motion Prior based Approach

Renhao Huang, Anthony Tompkins, Maurice Pagnucco, and 1 more author

In Conference on Lifelong Learning Agents, 2023

[ Bib ] [ Paper ]

@inproceedings{mp-traj,
  author = {Huang, Renhao and Tompkins, Anthony and Pagnucco, Maurice and Song, Yang},
  title = {{Towards Single Source Domain Generalisation in Trajectory Prediction: A Motion Prior based Approach}},
  booktitle = {Conference on Lifelong Learning Agents},
  year = {2023},
  pages = {227--243},
}