I’m a 3rd-year M.Phil. student in Electronic and Information Engineering at SIGS, Tsinghua University(THU), supervised by Prof. Lu Fang and Prof. Ruiqi Huang. Prior to that, I spent wonderful 4 years at Southeast University(SEU), advised by Prof. Yangang Wang and obtained my bachelor’s degree in 2022.

My research interests mainly lie in 2D and 3D computer vision, especially in:

• Visual Cognition: Human-centric Visual Analysis in Complex Scenes with Gigapixel Resolution.
• Embodied AI: Perceive, Comprehend, and Interact with the Environment through Physical Entities.

🔥 News

• 2024.03:  🎉🎉 Our paper DynamicTrack is accepted by ICME 2024!
• 2024.02:  🎉🎉 Our paper GigaTraj is accepted by CVPR 2024!

📖 Educations

• 2022.08 - 2025.06, M.Sc. in Electronic Information Engineering, Tsinghua University (THU), Beijing, China.
• 2018.08 - 2022.06, B.Sc. in Automation, Southeast University (SEU), Nanjing, China.

📝 Publications

ICME 2024

DynamicTrack: Advancing Gigapixel Tracking in Crowded Scenes

Yunqi Zhao^*, Yuchen Guo^*, Zheng Cao, Kai Ni, Ruqi Huang, Fu Fang^†

The IEEE International Conference on Multimedia & Expo (ICME), 2024.

Paper / Project Page

CVPR 2024

GigaTraj: Predicting Long-term Trajectories of Hundreds of Pedestrians in Gigapixel Complex Scenes

Haozhe Lin^*, Chunyu Wei^*, Li he^*, Yuchen Guo^*, Yunqi Zhao, Shanglong Li, Fu Fang^†

Conference on Computer Vision and Pattern Recognition (CVPR), 2024.

Paper / Dataset

ICIG 2021

Occluded Animal Shape and Pose Estimation from a Single Color Image

Yiming Xie, Yunqi Zhao, Shijian Jiang, Jiangyong Hu, Yangang Wang^†

International Conference on Image and Graphics, (ICIG), 2021.

🎖 Honors and Awards

• 2022, Excellent Graduate, Southeast University.
• 2021, Outstanding students, Southeast University.
• 2019, 2020, National Encouragement Scholarship, the Minister of Education, China.

🖥️ Research Experiences

• Pedestrian Trajectory Prediction in Gigapixel Complex Scenes 2023.07 – 2024.11
  Propose a trajectory prediction dataset featuring hundreds of pedestrians in gigapixel complex scenes.
  Generate pedestrian trajectories, conduct data annotation, and perform statistical analysis.
  Mentor: Prof. Lu Fang

• Multi-object Tracking in Gigapixel Complex Scenes 2022.09 – 2023.06
  Propose a contrastive learning-based detector for simultaneous head and body detection.
  Design a hierarchical association algorithm to utilize head and body cues for multi-object tracking.
  Mentor: Prof. Lu Fang and Prof. Ruqi Huang

💡 Projects

• Interactive General Object Grasping (Internship Program) 2024.04 – 2024.07
  Develop an interactive general object grasping system for robotic applications in open-world.
  Employ SAM for interactive target localization and GraspNet for general object grasping.
  

• Scene Reconstruction and Understanding for Intelligent Interaction (B.S. Thesis) 2021.11 – 2022.08
  Design a real-time system for indoor scene reconstruction, segmentation, and simplification.
  Develop local and global interaction patterns to improve 3D perception for visually impaired people.
  

• 3D Reconstruction of Animals from Single Color Images 2020.10 - 2021.05
  Employ a parametric model SMAL for 3D reconstruction of animals from a single color image.
  Predict the shape and pose parameters using the skeletal and contour information of animals.