Xinxuan Lu

About

I’m a third-year PhD Student in Computer Science at UCI, advised by Prof. Alex Berg & Prof. Charless Fowlkes.

I obtained my BS degree in Mathematics and Computer Science at Vanderbilt University. During my undergraduate study, I worked with Prof. Dan Work on 3D Vehicle Detection Models.

Email

LinkedIn

GitHub

CV

Research

I work on computer vision and generative models, with an emphasis on controllable image generation and 3D vision. My research focuses on developing more reliable and accurate text and 3D control for image generation. I aim to build unified latent spaces that integrate 2D images, 3D geometry, and text semantics.

Publications & Preprints

	Camera Control for Text-to-Image Generation via Learning Viewpoint Tokens Xinxuan Lu, Alex Berg, Charless Fowlkes Under review | Project Page
	Polygon Intersection-over-Union Loss for Viewpoint-Agnostic Monocular 3D Vehicle Detection Xinxuan Lu, Derek Gloudemans, Shuofeng Xia, Dan Work arXiv 2023