Xingyi Yang DSAI @ PolyU
The Hong Kong Polytechnic University Logo

Openings for PhD, Postdoc, MPhil, and RA @ PolyU DSAI

Supported by the Presidential Young Scholar Scheme

About Me

I am Xingyi Yang. I am now tenure-track Assistant Professor at The Hong Kong Polytechnic University (PolyU) in the Department of Data Science and Artificial Intelligence (DSAI), supported by the Presidential Young Scholar Scheme.

I am establishing a new research group dedicated to cutting-edge AI, with a focus on Generative AI, Computer Vision, Spatial Intelligence, World Models, and Efficient AI. I am looking for passionate Ph.D. students for 2027 Spring and Fall admission, as well as visiting students, RAs, and Postdocs.

Educational Background

I completed my Ph.D. at NUS, advised by Prof. Xinchao Wang. During my PhD, I was a visiting researcher at the University of Oxford, collaborating with Prof. Philip Torr (Fellow of the Royal Academy of Engineering). I received my B.Eng. from Southeast University and my M.S. from UC San Diego.

Research & Honors

My research focuses on Generative AI, Computer Vision, Spatial Intelligence, World Models, and Efficient AI, particularly their application to 3D/4D vision and physical world modeling. My work has been published in top-tier venues, including NeurIPS, ICML, ICLR, CVPR, ICCV, and ECCV, and has received several honors, such as the AAAI 2026 New Faculty Highlight, NeurIPS 2022 Best Paper Honorable Mention, Forbes 30 Under 30 Asia, and the 2023 Chinese Government Award for Outstanding Self-Financed Students Abroad.

💡 Research Interests

Our research aims to tackle fundamental challenges in AI by building smarter and more efficient systems that can understand and interact with the physical world. Key directions include:

  • Generative AI: Investigating structured generative models (e.g., Diffusion Models, LLMs) for controllable creation, content generation, scene modeling, and understanding.
  • Computer Vision & Spatial Intelligence: Fusing images, videos, and 3D/4D data to build intelligent systems that can understand, reconstruct, and reason about the physical world.
  • Efficient AI: Focusing on the computational efficiency, optimization, and adaptation of large pre-trained models.
  • World Models & Embodied AI: Exploring intelligent systems that can generate, reconstruct, reason about, and interact with dynamic physical worlds.

These areas are interrelated: 🤖 generative models provide general creative capabilities, 🌐 vision and spatial intelligence enhance understanding of the real world, ⚡ efficient AI ensures practical deployment, and 🧠 world models connect perception, generation, reasoning, and action.

I am always open to discussing new and exciting research ideas.

✨ What We Offer

  • Close and Consistent Mentorship: I am committed to working closely with each student, providing guidance and support to help them become independent and innovative researchers.
  • Full Financial and Resource Support: We provide stable research funding, access to computational resources (including a GPU cluster), and support for attending international conferences.
  • Collaboration and Networking Opportunities: I actively connect students with collaboration and internship opportunities at top institutions (e.g., NUS, Oxford, UCSD) and leading companies (e.g., Huawei, Meta, ByteDance).

📝 Who We Are Looking For

We are looking for highly motivated candidates who are passionate about research and have strong analytical and problem-solving skills. Ideal candidates will have:

  • A strong interest in one or more of our research areas and a desire to tackle challenging problems.
  • A solid background in CS, AI, EE, Mathematics, Statistics, or related fields.
  • Strong programming skills and proficiency in English.
  • For Postdoc Applicants: A Ph.D. degree (or will be receiving one soon) with a strong publication record.
  • For PhD Applicants: Prior research experience and publications are a plus.

I value diverse skill sets and intellectual curiosity beyond a narrow specialty. Feel free to highlight any other strengths in your application.

🗓️ Timelines & How to Apply

I plan to recruit 2–3 Ph.D. students for 2027 Spring and Fall admission. The deadline for Spring admission is September 30th. Outstanding candidates are strongly encouraged to apply for the PolyU President's PhD Fellowship Scheme (PPPFS) and the Hong Kong PhD Fellowship Scheme (HKPFS), for which I will provide full support.

I am also looking for 1–2 Postdocs and RAs. These positions are open year-round with competitive salaries and flexible start dates.

Research internship opportunities are available for undergraduate and master's students, primarily for Winter 2026 and Summer 2027.

To apply, please send your CV and any relevant materials (e.g., publications, portfolio) to xingyi.yang@polyu.edu.hk. Use the subject line: "[Position] Application - [Your Name]" (e.g., "PhD Application - Jane Doe"). Please include a brief statement of interest in your email.

✍️ A Final Word

Thank you for your interest. Choosing a supervisor is a significant decision, and for me, building this team is a new and exciting journey. My goal is to foster an open, collaborative, and supportive environment where we can explore new ideas and grow together.

As I am in the early stages of forming the group, I may not be able to reply to every email immediately. However, I assure you that I will review every application carefully.

I look forward to hearing from you and hope we can turn exciting ideas into reality together.