Carousel imageCarousel imageCarousel image

Dr. Xucong Zhang

Assistant Professor
Computer Vision Lab
Faculty of Electrical Engineering, Mathematics and Computer Science
Delft University of Technology
(TU Delft)

E-Mail: xucong.zhang [AT] tudelft.nl

Office: Room 5.W.620 (5th floor west)
Van Mourik Broekmanweg 6 (building 28)
2628 XE Delft

The Netherlands

Links: Google Scholar, CV [PDF], DBLP, LinkedIn, Twitter

Research Interest

My current research is about computer vision, machine learning, and human-computer interactions. My core research interest is human-centred computing as developing techniques for sensing, understanding, and serving the human user.

  • Gaze estimation

The task is the estimation of where the people looking from the input eye/face image. The methodology is mainly about computer vision and machine learning techniques such as CNN, GAN, VAE, domain adaptation, and attention mechanism. Gaze estimation has variant applications for human-computer interaction, human-robot interaction, VR/AR systems, psychology, and car driving. I have been working on the gaze estimation task for years.

  • Human-computer interaction

My current research on HCI is applying the developed gaze estimation method to perform the HCI task such as attention estimation, UI element optimization, and eye contact detection. This line of research is an important part of my long-term research goal. I am interested in HCI research and have been working on this area for years.

  • Human digitalization

My next five years (2022-2027) research plan is to develop the talking virtual avatar that can act like a human being. It would be a combination of multi-modality such as the facial landmark, hand pose, body pose, eye gaze, and speech. The research content includes new dataset collection with the state-of-the-art multi-camera capture system, and advanced deep learning techniques to drive the avatar. I gain rich experience from my previous works for these research contents. The generated virtual avatar can be used for HCI application and human-related modelling such as biomechanical engineering.

  • Biomechanical prior model

My next ten years (2022 - 2032) research plan is to develop computer vision and machine learning tools for biology- and healthcare-related research. As the starting point, I will introduce the biomechanical prior for the human-related models. I aim to provide novel face and body models with more accurate representation and require fewer data to train than the current skeleton- or mesh-based models.

Short Bio

I am an assistant professor at TU Delft since August 2021. I was a postdoc researcher (2018-2021) in the Advanced Interaction Technologies Lab at ETH Zurich, led by Prof. Otmar Hilliges. I did my PhD research (2013-2018) at Max Planck Institute for Informatics under the supervision of Prof. Andreas Bulling, and I received the doctoral degree ( summa cum laude) from Saarland University, Germany.

Latest news

[01.2022] We will organize the GAZE workshop at CVPR 2022.

[11.2021] One book chapter about gaze estimation and its applications is published in the book Artificial Intelligence for Human Computer Interaction: A Modern Approach.

[09.2021] Received the Facebook Research Award with Guohao Lan to design privacy-preserving eye-tracking for AR/VR and beyond. Thank you, Facebook!

[07.2021] One paper is accepted at ICCV 2021. See you online!

[01.2021] I will join TU Delft as an assistant professor in August 2021. One PhD position from me will be available soon. Very much looking forward to the new life in Delft.

[12.2020] We are going to organize "GAZE" workshop at CVPR 2021.

[09.2020] One paper is accepted at NeurIPS 2020. See you online!

[07.2020] One paper is accepted at BMVC 2020. See you online!

[05.2020] Two papers are accepted at ECCV 2020. See you online!

[04.2020] One paper is accepted by TIP.

[01.2020] We are going to organize "OpenEyes" workshop at ECCV 2020.

[09.2019] We won the OpenEDS Challenge 2019 (Synthetic Eye Generation Challenge).

[07.2019] One paper is accepted at ICCV 2019.