Zicong Fan

Zicong Alex Fan

Ph.D. student
Advanced Interactive Technologies Lab, ETH Zürich

E-Mail
zicong.fan@inf.ethz.ch
Address
Stampfenbachstrasse 48, 8092 Zürich, Switzerland
Room
ETH Zurich, Department of Computer Science, STD, G 23



Biography

I am a doctoral student at ETH supervised by Professor Otmar Hilliges, Professor Siyu Tang and Professor Michael J. Black. Before starting my Ph.D., I finished my B.Sc. (2018) and M.Sc. (2020) in Computer Science at The University of British Columbia, Vancouver, Canada. In my Masters, I worked in vision + language problems (e.g., visual grounding, visual commonsense reasoning) with Professor Leonid Sigal and Professor Jim Little.

Research Interests

My main research interests lie in computer vision and machine learning. These include topics such as hand pose estimation, human pose estimation and modelling the interaction of human-objects and human-scene. I am also interested in egocentric computer vision.

Research Opportunity for Prospective Students

Introduction: 3D hand pose/shape estimation and the modeling of human/object interaction is a pathway to real-world interactive AR/VR applications such as Microsofts HoloLens and Facebooks Occulus Rift. The reason is that we interact with the world using our hands to use tools and to socialize with others.

Research Directions: My main research interests lie in computer vision and machine learning, including but not limited to topics such as hand pose estimation, human pose estimation, and modeling the interaction of human-objects and human-scene. Currently, I offer two research streams: 1) Estimating 3d hand pose/shape from images; 2) Modelling the interaction of humans and objects. A stream-1 project could be devising a semi-supervised method for hand pose/shape estimation using temporal constraints. A stream-2 project could be creating a new task for the newly released GRAB dataset, which captures human-object interactions.

Requirements: I am looking for independent and highly motivated students who should have taken a recognized deep learning or a modern computer vision course (e.g., Machine Perception) and should be skilled in Python and PyTorch. The projects are research-oriented, and I encourage students to submit to top-tier computer vision conferences. I work closely with students during their projects as it is a great way for both of us to learn, and I would love to provide feedback and assistance for their projects.

Should you be interested in these areas or would like to know more, do not hesitate to contact me. P.S. students are encouraged (but not required) to bring their own ideas.

Keywords: computer vision, VR/AR, 3D reconstruction, 3D pose estimation, machine learning, neural networks, human-object interactions