Longfei Liu

News

• 02 / 2026
  FSOD-VFM has been accepted by ICLR 2026.
• 09 / 2025
  We have released our paper: Real-Time Object Detection Meets DINOv3 on arXiv. We not only achieve state-of-the-art results on the COCO dataset, but also, to the best of our knowledge, our S model is the first to surpass 50 AP(50.9) with fewer than 10M parameters.
• 09 / 2024
  We have released our paper: From COCO to COCO-FP: A Deep Dive into Background False Positives for COCO Detectors on arXiv. Our proposed COCO-FP dataset is now available here.

Research

^* indicates equal contribution.
⁺ indicates corresponding author.
Representative papers are highlighted.

FSOD-VFM: Few-Shot Object Detection with Vision Foundation Models and Graph Diffusion

Chen-Bin Feng^*, Youyang Sha^*, Longfei Liu, Yongjun Yu, Chi Man Vong⁺, Xuanlong Yu⁺, Xi Shen⁺

ICLR 2026

[Code]

Real-Time Object Detection Meets DINOv3

Shihua Huang^*, Yongjie Hou^*, Longfei Liu^*, Xuanlong Yu, Xi Shen⁺

Arxiv 2025

[Code]

From COCO to COCO-FP: A Deep Dive into Background False Positives for COCO Detectors

Longfei Liu, Wen Guo, Shihua Huang, Cheng Li, Xi Shen⁺

Arxiv 2024

[Code]

Others

• 07 / 2025: Join Intellindust as an algorithm engineer and continue working with Dr. Xi Shen.
• 09 / 2024: I am currently an intern at Intellindust, supervised by Dr. Xi Shen and Dr. Cheng Li.