Qingtao Yu (于庆涛)

I am currently a 2nd year PhD student at the Australian National University on a full scholarship, working under the supervision of Dylan Campbell, Hongdong Li and Tongliang Liu. Previously, I completed my Bachelor of Advanced Computing with First Class Honours at the Australian National University. My research focuses on generative models for computer vision. I grew up in Dalian, Liaoning, China.

Selected Papers

• TTSnap: Test-Time Scaling of Diffusion Models via Noise-Aware Pruning (Preprint 2025)
  Qingtao Yu, Changlin Song, Minghao Sun, Zhengyang Yu, Vinay Kumar Verma, Soumya Roy, Sumit Negi, Hongdong Li, Dylan Campbell
  


• Probability density geodesics in image diffusion latent space (CVPR 2025)
  Qingtao Yu, Jaskrit Singh, Zhaoyuan Yang, Peter Tu, Jing Zhang, Hongdong Li, Richard Hartely, Dylan Campbell
  


• When 3d bounding-box meets sam: Point cloud instance segmentation with weak-and-noisy supervision (WACV 2024)
  Qingtao Yu, Heming Du, Chen Liu, Xin Yu
  


• Benchmarking audio visual segmentation for long-untrimmed videos (CVPR 2024)
  Chen Liu, Peike Patrick Li, Qingtao Yu, Hongwei Sheng, Dadong Wang, Lincheng Li, Xin Yu
  

Professional Experience

• Academic Tutor at the Australian National University
  Teaching courses: Statistical Machine Learning, Computer Graphics, Computer Vision
• Serve as a reviewer for DICTA 2025, AJCAi 2025, RA-L