Hi, I’m Zhang Anqi, a Ph.D. student at LaVi Lab, The Chinese University of Hong Kong, advised by Prof. Liwei Wang. I received both my Bachelor’s (B.Eng., 2019-2023) and Master’s (M.Eng., 2023-2026) degrees from Beijing Institute of Technology, previously advised by Prof. Guangyu Gao.

My research interests mainly focus on Computer Vision and Multimodal Large Language Models. My past work includes Incremental/Few-shot Semantic Segmentation and Referring Segmentation.

🔥 News

  • 2026.02: Rethinking MLLM Itself as a Segmenter with a Single Segmentation Token is accepted by CVPR 2026
  • 2025.02: CoMBO: Conflict Mitigation via Branched Optimization for Class Incremental Segmentation is accepted by CVPR 2025
  • 2024.10: Bridge the Points: Graph-based Few-shot Segment Anything Semantically is accepted by NeurIPS 2024
  • 2024.07: Background Adaptation with Residual Modeling for Exemplar-Free Class-Incremental Semantic Segmentation is accepted by ECCV 2024

📝 Publications

CVPR 2026
paper

Rethinking MLLM Itself as a Segmenter with a Single Segmentation Token

Anqi Zhang, X Ji, Guangyu Gao, Jianbo Jiao, CH Liu, Y Wei

CVPR 2025
paper
NeurIPS 2024
paper

Bridge the Points: Graph-based Few-shot Segment Anything Semantically

Anqi Zhang, Guangyu Gao, Jianbo Jiao, CH Liu, Y Wei

🎖 Honors and Awards

  • National Scholarships x 2

📖 Educations

  • 2026.8 - Present, Ph.D. student, LaVi Lab, The Chinese University of Hong Kong (Prof. Liwei Wang)
  • 2025.5 - 2025.10, Visiting Research Student, University of Birmingham (Prof. Jianbo Jiao)
  • 2023.9 - 2026.6, Master’s Degree (M.Eng.), Beijing Institute of Technology (Prof. Guangyu Gao)
  • 2019.9 - 2023.6, Bachelor’s Degree (B.Eng.), Xuteli College (Beijing Institute of Technology)

💻 Internships

  • 2025.1 - 2025.4, Vision Algorithm Developer, Jingdong Retail