Yadan Luo

I am a Senior Lecturer and ARC Discovery Early Career Researcher Awardee (DECRA) at the School of Electrical Engineering and Computer Science, The University of Queensland (UQ), Australia. My research lies at the intersection of robust machine learning and open-world 3D vision, with a particular focus on test-time generalization under imperfect or inexact data conditions. I aim to develop learning systems that are not only theoretically grounded but also capable of delivering reliable performance in dynamic, real-world environments.

I received my PhD in Computer Science from UQ in 2021, supervised by Prof Helen Huang and Dr Mahsa Baktahmotlagh, following a Bachelor of Computer Science with Honours from the Yingcai Honors College at UESTC. My academic trajectory has been recognized through several competitive awards, including the Google PhD Fellowship in Machine Learning, UQ Foundation Research Excellence Award, the ICT Young Achiever Award, and a few Best Paper Awards. In parallel to my research, I contribute extensively to the academic and professional community. I currently serve as Associate Editor for IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), Area Chair for CVPR/NeurIPS/ICLR/ICML and ACM Multimedia, and organizing committee member for ACM Multimedia (MM’24) and The Web Conference 2025 (WWW’25). I am also actively involved in collaborative research projects with industry and government, translating core machine learning innovations into domains such as digital manufacturing, road infrastructure, and intelligent transportation systems.

I am recruiting PhDs in computer vision and machine learning. Please feel free to drop me an email with your CV. Scholarships may be available for the 2026 enrollment.

News

Jun, 2025	Excited to share that our latest paper on test-time adaptation for CLIP/SigLIP series VLMs has been accepted for presentation at ICCV 2025! Our method requires no additional training, and achieves comparable performance at just 70% of the original computational cost. Check out the details in the paper!
Jun, 2025	My conference memo of CVPR’25 is now available! Read it here. This memo provides an overview of workshops and tutorials, oral sessions, and a curated selection of posters (coming soon!) in 3DV and autonomous driving. Attending CVPR in Nashville was an inspiring and memorable experience—grateful for the opportunity to connect with new friends in this field :)
May, 2025	I will serve as an Area Chair for NeurIPS 2025.
Feb, 2025	One end-to-end autonomous driving paper has been accepted for presentation at CVPR 2025.
Feb, 2025	ICLR 2025 has accepted one Poster and one Oral paper (1.8%). Congratulations to the team on this remarkable achievement!
Jan, 2025	I will serve as an Area Chair for ICML 2025.
Nov, 2024	Grateful to receive both the Outstanding Service Award and the Outstanding Area Chair Award at ACM MM 2024. Thank you to the community for the recognition and support! 🌟
Oct, 2024	I will serve as an Area Chair for CVPR 2025.
Oct, 2024	I will serve as an Area Chair for ICLR 2025.
Oct, 2024	3 papers on prompt expansion for class-agnostic detection, dataset distillation and cross-modal retrieval with incomplete labels has been accepted by NeurIPS 2024!

Selected Publications

Is Less More? Exploring Token Condensation as Training-free Adaptation for CLIP

Zixin Wang, Dong Gong, Sen Wang, Zi Huang, and 1 more author

In IEEE/CVF International Conference on Computer Vision, (ICCV) 2025

ICCV

arXiv Bib PDF Code

@inproceedings{DBLP:conf/cvpr/momad,
  author = {Wang, Zixin and Gong, Dong and Wang, Sen and Huang, Zi and Luo, Yadan},
  title = {Is Less More? Exploring Token Condensation as Training-free Adaptation for CLIP},
  booktitle = {{IEEE/CVF} International Conference on Computer Vision, {(ICCV)}},
  year = {2025},
}

Don’t Shake the Wheel: Momentum-Aware Planning in End-to-End Autonomous Driving

Ziying Song, Caiyan Jia, Lin Liu, Hongyu Pan, and 6 more authors

In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2025

CVPR

arXiv Bib PDF Code

@inproceedings{DBLP:conf/cvpr/momae,
  author = {Song, Ziying and Jia, Caiyan and Liu, Lin and Pan, Hongyu and Zhang, Yongchang and Wang, Junming and Zhang, Xingyu and Xu, Shaoqing and Yang, Lei and Luo, Yadan},
  title = {Don't Shake the Wheel: Momentum-Aware Planning in End-to-End Autonomous Driving},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year = {2025},
}

MOS: Model Synergy for Test-Time Adaptation on LiDAR-Based 3D Object Detection

Zhuoxiao Chen, Junjie Meng, Mahsa Baktashmotlagh, Yonggang Zhang, and 2 more authors

In International Conference on Learning Representations (ICLR), **Oral Presentation** 2025

ICLR

arXiv Bib PDF Code

@inproceedings{DBLP:conf/iclr/mos,
  author = {Chen, Zhuoxiao and Meng, Junjie and Baktashmotlagh, Mahsa and Zhang, Yonggang and Huang, Zi and Luo, Yadan},
  title = {MOS: Model Synergy for Test-Time Adaptation on LiDAR-Based 3D Object Detection},
  booktitle = {International Conference on Learning Representations ({ICLR}), **Oral Presentation**},
  publisher = {OpenReview.net},
  year = {2025},
}

PolaFormer: Polarity-aware Linear Attention for Vision Transformers

Weikang Meng, Yadan Luo, Xin Li, Dongmei Jiang, and 1 more author

In International Conference on Learning Representations (ICLR) 2025

ICLR

arXiv Bib PDF Code

@inproceedings{DBLP:conf/iclr/polaformer,
  author = {Meng, Weikang and Luo, Yadan and Li, Xin and Jiang, Dongmei and Zhang, Zheng},
  title = {PolaFormer: Polarity-aware Linear Attention for Vision Transformers},
  booktitle = {International Conference on Learning Representations ({ICLR})},
  publisher = {OpenReview.net},
  year = {2025},
}

ConjNorm: Tractable Density Estimation for Out-of-Distribution Detection

Bo Peng*, Yadan Luo*, Yonggang Zhang, Yixuan Li, and 1 more author

In International Conference on Learning Representations (ICLR) 2024

ICLR

arXiv Bib

@inproceedings{DBLP:conf/iclr/conjnorm,
  author = {Peng*, Bo and Luo*, Yadan and Zhang, Yonggang and Li, Yixuan and Fang, Zhen},
  title = {ConjNorm: Tractable Density Estimation for Out-of-Distribution Detection},
  booktitle = {International Conference on Learning Representations (ICLR)},
  publisher = {OpenReview.net},
  year = {2024},
}