Jingyi Xu

About

Hi! I am Jingyi Xu. I received my Ph.D. degree at Stony Brook University in 2024, working with Professor Dimitris Samaras. Before that, I received my B.S. degree at Nankai University in 2019. My research interets lie in the field of Computer Vision and Deep Learning, with a focus on generative modeling and representation learning.

Updates

Jun 2025: Three ICCV’25 papers on token merging, gaze target estimation and object counting are accepted.
May 2025: One MICCAI’25 paper on 3D human heart modeling is accepted.
Feb 2025: One CVPR’25 paper on personalized scanpath prediction is accepted.
Jan 2025: One ICLR’25 paper on compositional text-to-image generation is accepted as spotlight.
Oct 2024: One WACV’25 paper on object counting is accepted.
Jul 2024: One ECCV’24 paper on sample quality assessment is accepted.
Jun 2024: Successfully defended my Ph.D.!
May 2023: Start my internship at Amazon (Seattle).
Feb 2023: Two CVPR’23 papers on object detection and object counting are accepted.
Mar 2022: One CVPR’22 paper on image classification is accepted.
Jul 2021: One ICCV’21 paper on fine-grained few-shot image classification is accepted.
May 2021: Start my internship at NEC Laboratories America.
Sep 2019: Start my Ph.D. at Stony Brook Computer Science!

Publications

Importance-Based Token Merging for Efficient Image and Video Generation.
Haoyu Wu, Jingyi Xu, Hieu Le, Dimitris Samaras.
IEEE International Conference on Computer Vision (ICCV) 2025.
Multi-view Gaze Target Estimation.
Qiaomu Miao, Vivek Raju Golani, Jingyi Xu, Progga Paromita Dutta, Minh Hoai, Dimitris Samaras.
IEEE International Conference on Computer Vision (ICCV) 2025.
Counting Stacked Objects.
Corentin Dumery, Noa Ette, Aoxiang Fan, Ren Li, Jingyi Xu, Hieu Le, Pascal Fua.
IEEE International Conference on Computer Vision (ICCV) 2025.
Pairwise-Constrained Implicit Functions for 3D Human Heart Modeling.
Hieu Le, Jingyi Xu, Nicolas Talabot, Jiancheng Yang, Pascal Fua.
Medical Image Computing and Computer Assisted Intervention (MICCAI) 2025.
Few-shot Personalized Scanpath Prediction.
Ruoyu Xue, Jingyi Xu, Sounak Mondal, Hieu Le, Greg Zelinsky, Minh Hoai , Dimitris Samaras.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2025.
Enhancing Compositional Text-to-Image Generation with Reliable Random Seeds.
Shuangqi Li, Hieu Le, Jingyi Xu, Mathieu Salzmann.
International Conference on Learning Representations (ICLR) 2025 (Spotlight).
Learning to Count from Pseudo-Labeled Segmentation.
Jingyi Xu, Hieu Le, Dimitris Samaras.
European Conference on Computer Vision (WACV) 2025.
Assessing Sample Quality via the Latent Space of Generative Models.
Jingyi Xu, Hieu Le, Dimitris Samaras.
European Conference on Computer Vision (ECCV) 2024.
Zero-Shot Object Counting.
Jingyi Xu, Hieu Le, Vu Nguyen, Viresh Ranjan, Dimitris Samaras.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2023.
Generating Features With Increased Crop-related Diversity For Few-shot Object Detection.
Jingyi Xu, Hieu Le, Dimitris Samaras.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2023.
Generating Representative Samples for Few-Shot Classification.
Jingyi Xu, Hieu Le
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2022.
Variational Feature Disentangling for Fine-Grained Few-Shot Classification.
Jingyi Xu, Hieu Le, Mingzhen Huang, ShahRukh Athar, Dimitris Samaras.
IEEE International Conference on Computer Vision (ICCV) 2021.
RegularFace: Deep Face Recognition via Exclusive Regularization.
Kai Zhao, Jingyi Xu, Ming-Ming Cheng.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.

Experience

Research Intern at Amazon, Seattle, May 2023 - July 2023
Research Intern at NEC Lab, CA, May 2021 - Aug 2021
Research Intern at SenseTime, Beijing, Dec 2018 - July 2019
Research Intern at Panasonic R&D Center,Singapore, Sep 2018 - Dec 2018