Hi there Welcome to my Homepage!

I’m a Ph.D. student at MMT lab, University of Liverpool, co-advised by Dr. Siyue Yu, Prof. Jimin Xiao and Dr. Yi Dong. I’m also fortunate to work closely with Dr. Mingjie Sun from Soochow University and Dr. Bingfeng Zhang from China University of Petroleum (East China).

Previously, I worked as a camera engineer at vivo Mobile Communication Ltd, focusing on auto-focus and image stabilization. I graduated from MePro led by Prof. Yao Zhao (IEEE Fellow), at Beijing Jiaotong University, advised by Prof. Meiqin Liu and Prof. Chao Yao.

My research focuses on open-vocabulary dense perception, multimodal visual grounding, concept/reasoning/text-guided segmentation and multimodal large language models, with Google Scholar citations.

News

2026.07 🎉🎉 One regular paper accepted by ECCV 2026.
2026.05 🎉🎉 One regular paper accepted by ICML 2026.
2026.03 🎉🎉 One regular paper and one Findings paper accepted by CVPR 2026.
2025.08 🎉🎉 SFP selected as an ICCV 2025 Highlight paper. Many thanks to all co-authors!
2025.06 🎉🎉 One regular paper accepted by ICCV 2025.
2023.10 🎉🎉 Two regular papers accepted by ACM MM 2023.
2023.04 🎉🎉 One regular paper accepted by IEEE T-CSVT.

Experience

University of Liverpool
2024.09 - Present
Ph.D. in Computer Science
MMT lab, co-advised by Dr. Siyue Yu, Prof. Jimin Xiao, and Dr. Yi Dong.

vivo Mobile Communication Ltd.
2023.07 - 2024.08
Camera Engineer
Auto-focus and image stabilization.

Beijing Jiaotong University
2021.09 - 2023.06
M.Eng. in Electronic Engineering
MePro, advised by Prof. Meiqin Liu and Prof. Chao Yao.

Publications

(* equal contribution · † corresponding author · ‡ project leader)

Don't Starve the Boundaries: Boundary-Constrained Label Propagation for Weakly Supervised 3D Segmentation
Shuwei Wu, Shuo Jin, Zhijin He, Siyue Yu, ENG GEE LIM, Qiufeng Wang, Jimin Xiao.
Boundary-aware pseudo-label propagation for weakly supervised point-cloud segmentation.
ECCV 2026 [paper] [project] [code] [poster]

VIP: Visual-guided Prompt Evolution for Efficient Dense Vision-Language Inference
Hao Zhu*, Shuo Jin*, Wenbin Liao, Jiayu Xiao, Yan Zhu, Siyue Yu, Feng Dai.
Visually guided prompt evolution for efficient dense vision-language inference.
ICML 2026 [paper] [project] [code] [poster]

TF-SSD: A Strong Pipeline via Synergic Mask Filter for Training-free Co-salient Object Detection
Zhijin He*, Shuo Jin*, Siyue Yu, Shuwei Wu, Bingfeng Zhang, Li Yu, Jimin Xiao.
A training-free pipeline combining mask candidates, saliency, and cross-image semantic consistency.
CVPR 2026 [paper] [project] [code] [poster]

TALENT: Target-aware Efficient Tuning for Referring Image Segmentation
Shuo Jin, Siyue Yu, Bingfeng Zhang, Chao Yao, Meiqin Liu, Jimin Xiao.
Target-aware parameter-efficient tuning to ground the instance described by a referring expression.
CVPR 2026 Findings [paper] [project] [code] [poster]

Feature Purification Matters: Suppressing Outlier Propagation for Training-Free Open-Vocabulary Semantic Segmentation
Shuo Jin, Siyue Yu, Bingfeng Zhang, Mingjie Sun, Yi Dong, Jimin Xiao.
Suppressing attention outliers to improve spatial perception in open-vocabulary segmentation.
ICCV 2025 Highlight [paper] [project] [code] [poster]

Patents

A Co-salient Object Detection Algorithm Based on Democratic Prototype Feature Mining, CN119963800A.

Services

Reviewer for IEEE T-IP, ACM MM 2025, ICCV 2025, AAAI 2026/2027, CVPR 2025/2026, and ICLR 2027.

Teaching

INT305: Machine Learning (Undergraduate course at XJTLU), Teaching Assistant.
INT302: Image Processing (Undergraduate course at XJTLU), Teaching Assistant.