Ke Jin (金柯)

Email        Google Scholar        Github

I am a PhD student supervised by Qi Ye in Zhejiang University. Prior to that, I obtained my Bachelor’s degree from Southeast University in 2024.

My current research interest lies in 3D computer vision and Embodied AI, particularly in dexterous manipulation.

This work improves semi-dense local feature matching by introducing scale-aware coarse matching and flow-based fine refinement for better local consistency.

We propose a novel dataset named Varied Drone Dataset (VDD) for semantic segmentation

We propose a novel feature fusion module to make language-guided lightweight semantic segmentation practical.

The website template was adapted from Jon Barron.