Yuqun Wu

I am a second year PhD student in Computer Science department of University of Illinois at Urbana-Champaign, advised by Prof. Derek Hoiem. I also work closely with Prof. Shenlong Wang.

Prior to that, I also spent one year as a MS student and received my Bachelor of Science degree at UIUC, majoring in Computer Science & Statistics.

Email  /  CV  /  Linkedin

profile photo
Research

My research interests mainly lie in computer vision, especially for 3D scene understanding, video understanding, neural rendering, and geometry reconstruction.

clean-usnob SceneDiff: Geometric-Semantic Consistency for Multiview Change Detection
Yuqun Wu, Chih-hao Lin, Aditi Tiwari, Chuhang Zou, Shenlong Wang, Derek Hoiem
Under Review
Coming Soon

Propose a dataset and a new method for object change detection between a pair of captures (images or videos) of the same scene at different times.

clean-usnob TextRegion: Text-Aligned Region Tokens from Frozen Image-Text Models
Yao Xiao, Qiqian Fu, Heyi Tao, Yuqun Wu, Zhen Zhu, and Derek Hoiem
Under Review
Preprint , Code

Propose a training-free approach to create text compatible region tokens, enabling powerful zero-shot region-level understanding with existing image-text models.

clean-usnob MonoPatchNeRF: Improving Neural Radiance Fields with Patch-based Monocular Guidance
Yuqun Wu*, Jae Yong Lee*, Chuhang Zou, Shenlong Wang , and Derek Hoiem
3DV 2025
Project Page , Preprint , Code

Create 3D models that provide accurate geometry and view synthesis, partially closing the large geometric performance gap between NeRF and traditional MVS methods

clean-usnob Plenoptic PNG: Real-Time Neural Radiance Fields in 150 KB
Jae Yong Lee, Yuqun Wu, Chuhang Zou, Derek Hoiem , and Shenlong Wang
3DV 2025
Project Page , Preprint , Code

Encode 3D scenes into extremely compact representation from 2D images and enable its transmittance, decoding and rendering in real-time across various platforms via traditional GL pipeline.

clean-usnob Region-Based Representations Revisited
Michal Shlapentokh-Rothman*, Ansel Blume*, Yao Xiao, Yuqun Wu, Sethuraman T V, Heyi Tao, Jae Yong Lee, Wilfredo Torres, Yu-Xiong Wang, and Derek Hoiem
CVPR 2024
Project Page , Preprint , Code

Investigate region based representation, combining class-agnostic segmentation from SAM and dense features from foundation models, for a wide variety of tasks, including semantic segmentation, object-based image retrieval, and multi-image analysis.

clean-usnob Sparse SPN: Depth Completion from Sparse Keypoints
Yuqun Wu*, Jae Yong Lee*, and Derek Hoiem
Preprint

Propose a novel method that outperforms existing depth completion pipelines given sparse keypoint depth, and reconstructs complete point clouds given SfM setups

clean-usnob QFF: Quantized Fourier Features for Neural Field Representations
Jae Yong Lee, Yuqun Wu, Chuhang Zou, Shenlong Wang, and Derek Hoiem
Preprint

Present Quantized Fourier Features (QFF), which encodes features in bins of Fourier features, and can result in smaller model size, faster training, and better quality outputs for various applications of neural representation

Service

Teaching Assistant: CS 445: Computational Photography (Fall 2022), CS 441: Applied Machine Learning (Spring 2023)

Reviewer: CVPR 2025, ECCV 2024, WACV 2025

Updated on Sep 5 2025. Thanks Jon Barron for his amazing template.