Xi Chen

I am a first-year PhD student in Computer Science and Engineering at UC San Diego, advised by Prof. Henrik Christensen. Previously, I was fortunate to work with Prof. Hao Su. I completed my bachelor's and master's degrees at Zhejiang University, where I was advised by Prof. Xiaowei Zhou.

I am interested in AI systems that turn visual observations into models they can reason with and act upon. My previous work focuses on 3D reconstruction, view synthesis, and inverse rendering; I am now exploring robot learning and AI agents for automated research.

Publications

LVSPM Figure 1 comparing novel-view synthesis and predicted camera poses from 128 uncalibrated views

LVSPM: Long Sequence View Synthesis and Pose Estimation Model

Xi Chen, Yachi Zhang, Linghao Chen, Minghua Liu, Hao Su, Zexiang Xu, Xiaoshuai Zhang

ECCV 2026

TL;DR: Jointly estimates camera poses and synthesizes novel views from hundreds of uncalibrated input images.

Project Paper Code

EnvGS Figure 1 comparing reflections reconstructed by the method and prior work

EnvGS: Modeling View-Dependent Appearance with Environment Gaussian

Tao Xie^*, Xi Chen^*, Zhen Xu, Yiman Xie, Yudong Jin, Yujun Shen, Sida Peng, Hujun Bao, Xiaowei Zhou

CVPR 2025

TL;DR: Models reflections and other view-dependent effects with environment Gaussians.

Project Paper Code

IntrinsicAnything Figure 1 illustrating shading–albedo and shadow–albedo ambiguity

IntrinsicAnything: Learning Diffusion Priors for Inverse Rendering Under Unknown Illumination

Xi Chen, Sida Peng, Dongchen Yang, Yuan Liu, Bowen Pan, Chengfei Lv, Xiaowei Zhou

ECCV 2024

TL;DR: Uses diffusion priors to recover material properties under unknown illumination.

Project Paper Code

NeuralRecon dense indoor reconstruction with estimated camera trajectory

NeuralRecon: Real-Time Coherent 3D Scene Reconstruction From Monocular Video

Xi Chen, Jiaming Sun, Yiming Xie, Hujun Bao, Xiaowei Zhou

IEEE TPAMI 2024

TL;DR: Reconstructs coherent 3D scene geometry from monocular video in real time.

Paper Project Code

NeuralRecon-W landmark geometry and surface-normal reconstruction

Neural 3D Reconstruction in the Wild

Jiaming Sun, Xi Chen, Qianqian Wang, Zhengqi Li, Hadar Averbuch-Elor, Xiaowei Zhou, Noah Snavely

SIGGRAPH 2022

TL;DR: Reconstructs scene geometry from unconstrained internet photo collections.

Project Paper Code

^* indicates equal contribution.