Yuanfeng Ji (纪源丰)

github
email
linkedin
google-scholar
resume

About Me

As a final year Ph.D. student in the Department of Computer Science at the University of Hong Kong, I am conducting research under the supervision of Prof. Ping Luo and Prof. Wenping Wang (opens new window). My academic journey began with a foundation in EE from Shenzhen University, under the guidance of Prof. Lin Di.

My research interests are currently centered in the field of AI, with a special focus on its application in medicine, with the goal of developing systems that are not only powerful, but also trustworthy.

Inspired by pioneers in the field, my work seeks to contribute to the potential of AI to improve healthcare diagnostics and treatment strategies. The goal is to advance the intersection of technology and healthcare, enabling improved patient outcomes through the application of AI.

Please feel free to contact us via email at u3008013@connect.hku.hk.

News

  • [2024-01] AutoBench is accepted by ICLR24. See you in Wien
  • [2023-11] I joint the Prof. Ruijiang Li (opens new window)'s group at Stanford University as a visiting student.
  • [2023-07] DDP is accepted by ICCV23. See you in Paris
  • [2022-11] DrugOOD is accepted by AAAI23.
  • [2022-10] AMOS is accepted by NIPS22. See you in New Orleans

Publications

→ Full list (opens new window)

Large Language Models as Automated Aligners for benchmarking Vision-Language Models

Yuanfeng Ji*, Chongjian Ge*, Weikai Kong, Enze Xie, Zhengying Liu, Zhengguo Li, Ping Luo

ICLR 2024

Introduction: This research explores the potential of large language models as automated aligners, setting a new benchmark in vision-language model evaluation.

[Paper (opens new window)] [Code&Data(wip) (opens new window)]

SyNDock: N Rigid Protein Docking via Learnable Transformation Synchronization

Yuanfeng Ji, Yatao Bian, Guoji Fu, Peilin Zhao, Ping Luo

Tech report

Introduction: SyNDock presents an innovative approach to protein docking, utilizing learnable transformation synchronization for enhanced accuracy and efficiency.

[Paper (opens new window)] [Code&Data(wip)]

DDP: Diffusion Model for Dense Visual Prediction

Yuanfeng Ji*, Zhe Chen*, Enze Xie, Lanqing Hong, Xihui Liu, Zhaoqiang Liu, Tong Lu, Zhenguo Li, Ping Luo

ICCV 2023

Introduction: A groundbreaking approach to dense visual prediction, employing diffusion models to enhance accuracy and efficiency.

[Paper (opens new window)] [Code (opens new window)]

DrugOOD: Out-of-Distribution (OOD) Dataset Curator and Benchmark for AI-aided Drug Discovery

Yuanfeng Ji*, Lu Zhang*, Jiaxiang Wu, Bingzhe Wu, Lanqing Li, Long-Kai Huang, Tingyang Xu, Yu Rong, Jie Ren, Ding Xue, Houtim Lai, Wei Liu, Junzhou Huang, Shuigeng Zhou, Ping Luo, Peilin Zhao, Yatao Bian

AAAI 2023 (Oral)

Introduction: DrugOOD serves as a curator and benchmark for AI-driven drug discovery, focusing on affinity prediction problems with noise annotations.

[Paper (opens new window)] [Code (opens new window)] [Project (opens new window)]

AMOS: A Large-Scale Abdominal Multi-Organ Benchmark for Versatile Medical Image Segmentation

Yuanfeng Ji, Haotian Bai, Chongjian Ge, Jie Yang, Ye Zhu, Ruimao Zhang, Zhen Li, Lingyan Zhang, Wanling Ma, Xiang Wan, Ping Luo

Neuips 2022 (Oral)

Introduction: AMOS stands as a large-scale benchmark for abdominal multi-organ segmentation, paving the way for advancements in medical image analysis.

[Paper (opens new window)] [Code (opens new window)] [Challenge (opens new window)]

Multi-compound Transformer for Accurate Biomedical Image Segmentation

Yuanfeng Ji, Ruimao Zhang, Huijie Wang, Zhen Li, Lingyun Wu, Shaoting Zhang, Ping Luo

MICCAI 2021 (Early Accept)

Introduction: This work introduces a transformative approach in biomedical image segmentation, leveraging a multi-compound transformer architecture for enhanced accuracy.

[Paper (opens new window)] [Code (opens new window)]

UXNet: Searching Multi-level Feature Aggregation for 3D Medical Image Segmentation

Yuanfeng Ji, Ruimao Zhang, Zhen Li, Jiamin Ren, Shaoting Zhang, Ping Luo

MICCAI 2021 (Early Accept)

Introduction: UXNet propels the search for multi-level feature aggregation in 3D medical image segmentation through an AutoML tool for network design.

[Paper (opens new window)] [[Code(coming)]]

PRSNet: Part Relation and Selection Network For Bone Age Assessment

Yuanfeng Ji, Hao Chen, Dan Lin, Xiaohua Wu, Di Lin

MICCAI 2020 (Early Accept)

Introduction: PRSNet innovates bone age assessment by integrating part relation and selection networks to streamline the analysis process.

[Paper (opens new window)]

RANet: Region Attention Network for Semantic Segmentation

Dingguo Shen*, Yuanfeng Ji*, Ping Li, Yi Wang, Di Lin

Neuips 2020

Introduction: RANet leverages region-based attention mechanisms to enhance the performance of semantic segmentation tasks.

[Paper (opens new window)] [Code (opens new window)]

Multi-Scale Context Interwining for Semantic Segmentation

Di Lin, Yuanfeng Ji, Dani Lischinski, Daniel Cohen-Or, Hui Huang

ECCV 2018

Introduction: MSCI introduces an innovative approach to semantic segmentation by intertwining multi-scale contextual information, enhancing the accuracy and robustness of the segmentation process.

[Paper (opens new window)] [Project (opens new window)]

Challenges & Achievements

Professional Activities

Experience

I am deeply grateful for the growth and learning I have experienced under the guidance of my respected mentors.

2023-12-03: 12/3/2023, 11:23:25 AM