Yingshu CHEN (Susan) 陈颖舒

E-MAIL: yingshu2008[AT]gmail[DOT]com | LINKEDIN

Yingshu CHEN is currently a PhD candidate in the Department of Computer Science and Engineering, the Hong Kong University of Science and Technology (HKUST) supervised by Sai-Kit Yeung and Ajay Joneja. Her research interests cover computer vision and computer graphics for computational design such as 2D and 3D style transfer. Yingshu received her B.Eng in Digital Media Technology from Zhejiang University and M.Sc in Computer Science from The University of Hong Kong.

EDUCATION AND WORKING EXPERIENCE

August 2019 - Now, PhD student in Department of Computer Science and Engineering, HKUST

Summer 2019, Research Assistant of Division of Integrative Systems and Design, HKUST

2018 – 2019, Analytics Programmer, Aigens Technology Limited (Hong Kong)

2016 – 2018, Master of Science in Computer Science, The University of Hong Kong

2012 – 2016, Bachelor of Engineering in Digital Media Technology, Zhejiang University

SELECTED PUBLICATIONS

360VOT: A New Benchmark Dataset for Omnidirectional Visual Object Tracking

Huajian Huang, Yinzhe Xu, Yingshu Chen, Sai-Kit Yeung
International Conference on Computer Vision (ICCV) 2023.

[Project] [Arxiv] [Toolkit]

360VOT aims at the novel task of omnidirectional object tracking with new video object tracking benchmark dataset and evaluation metrics.

360Roam: Real-Time Indoor Roaming Using Geometry-Aware 360ᵒ Radiance Fields

Huajian Huang, Yingshu Chen, Tianjian Zhang, Sai-Kit Yeung

[Project] [Arxiv]

360Roam is a novel scene-level NeRF system that can synthesize novel views of large-scale indoor scenes and support real-time indoor roaming.

Neural Scene Decoration from a Single Photograph

Hong Wing Pang, Yingshu Chen, Phuoc-Hieu T. Le, Binh-Son Hua, Thanh Nguyen, Sai-Kit Yeung
European Conference on Computer Vision (ECCV) 2022.

[Project] [Paper] [Supp] [Arxiv] [Code]

Given a photograph of an empty indoor space and a list of decorations with layout determined by user, we aim to synthesize a new image of the same space with desired furnishing and decorations.

Time-of-Day Neural Style Transfer for Architectural Photographs

Yingshu Chen, Tuan-Anh Vu, Ka-Chun Shum, Binh-Son Hua, Sai-Kit Yeung
International Conference on Computational Photography (ICCP) 2022.

[Paper] [Project] [Code] [Data]

Architectural Photography Style Transfer aims to transfer background dynamic texture and chrominance, and transfer sufficient styles for foreground while keeping foreground geometry intact.

PROJECTS

3D Advertisement Poster (WebGL): A web-based demo for 3D advertisment display.

Human Computer Interaction via Brainwave Control (C++ and Qt): Demo Video

    - Study the effect of user interface (UI) elements and human-induced factors on electroencephalography (EEG) signals, proposing a brain-computer interface (BCI) typing system through EEG data retrieved from the wireless Emotiv neuroheadset.

Movie Trend Visualization: responsible for data collection, overall design, top movie visualization and analysis, webpage design.

Design and Method Study of Two-handed Natural Interactive Scene Modeling Applied on Virtual Reality (Unity with C#):

    - Free hand gesture control for simple modeling with Leap Motion.

Original Design of Websites with Interaction Visualization: UG Student Research Training Program (SRTP)

OTHER REFERENCES