Zhengjie (Jerry) Xu

I am a first-year PhD student in Computer Science and Engineering at the University of Michigan, Ann Arbor. I am luckily advised by Prof. Stella Yu. I am broadly interested in computer vision and robotics.

Previously, I obtained my M.S.E. in Robotics from University of Pennsylvania, and B.S. in Computer Engineering and Applied Mathematics from UCSD. During my master, I was fortunate to work with Prof. Jianbo Shi on image generation and reconstruction. I also worked with Prof. Hao Su and Prof. Garrison W. Cottrell during my undergrad.

Email  /  GitHub  /  Google Scholar

profile photo


project image

Perceptual Artifacts Localization for Image Synthesis Tasks

Lingzhi Zhang*, Zhengjie Xu*, Connelly Barnes, Yuqian Zhou, Qing Liu, He Zhang, Zhe Lin, Eli Shechtman, Sohrab Amirghodsi, Jianbo Shi
ICCV, 2023

Paper, code, and data comming soon! We generalize Perceptual Artifacts Localization to ten diverse image synthesis, and shows promising accuracy. We also show the effectiveness of automatic artifacts fixing and quality assessment as downstream applications.

project image

Optimizing Algorithms From Pairwise User Preferences

Lukas Zhornyak*, Zhengjie Xu*, Haoran Tang*, Jianbo Shi
arxiv, 2022
arxiv / code /

We present HashEncoding, a novel autoencoding architecture that leverages a non-parametric multiscale coordinate hash function to facilitate a per-pixel decoder without convolutions. By leveraging the space-folding behaviour of hashing functions, HashEncoding allows for an inherently multiscale embedding space that remains much smaller than the original image.

project image

Generalization in Cardiac Image Segmentation

Zhengjie Xu, Zixiang Zhou, Garrison W. Cottrell, and Mai H. Nguyen
IEEE International Conference on Big Data, 2021
arxiv /

We focus on the task of cardiac image semantic segmentation and synthesize five different real-world scenarios to find the optimal training approach for deep learning models to achieve good generalization across different datasets.

Other Projects

These include coursework, side projects and unpublished research work.

project image

Super-pixel Sampling in 3D

UCSD CSE 291: Deep Learning meets Geometry
paper /

We implemented a differentiable grouping module and combine it with off-the-shelf point cloud analysis frameworks (PointNet++). The module optimizes the embedding space to enforce the local feature consistency. The learnt mid-level abstraction can be used to discover geometry primitives as well as simplify the downstream tasks. With this module, we out-performed the original framework in point segmentation task.


I have a lovely ragdoll as you can see from my profile photo. His name is Dodo(代代) and he is now 2 years old!

I am a sports fan, and especially a soccer fan. My favorite teams are FC Barcelona and Arsenal. I also played a wide variety of sports, including soccer, basketball, tennis, golf, etc.

I started playing video games when I was only 3. The first game I played was Pokémon Gold on GBA and Pokémon is my all-time favorite. I am currently obssessed with Dota 2 and Apex Legends.

Design and source code from Jon Barron's website