Dian Zheng 郑典

I'm passionately obsessed with novelty, but at the same time, ideas are cheap, show me your code.

I am currently a third-year master's student at Sun Yat-sen University (SYSU), where I am advised by Wei-Shi Zheng. Before coming to SYSU, I obtained my B.E. degree at the Dalian University of Technology.

I'm fortunate to have internships at Alibaba and Shanghai AI Lab and I am honored to be advised by Ziwei Liu.

My main research focuses on AIGC and its applications on downstream tasks. In the following years, I will focus on exploring the potential of unifying visual generation and understanding, unleashing the power of both and making a step to the true world simulator. I am always open to research discussions and collaborations.

Email  / Google Scholar  /   Github

profile photo

News
[03/2025] VBench-2.0 is released! Test the boundaries of your methods!
[02/2025] 3 papers are accepted by CVPR2025 (3/4, 75% success rate).
[01/2025] DiffuVolume has finally been accepted by IJCV2025 after a long wait!
[12/2024] Awarded Xiaomi Grand Prize Scholarship of Sun Yat-Sen University.
[11/2024] Ended a wonderful 6-month internship at Alibaba.
[11/2024] SpatialDreamer is released, I would call it the most advanced spatial video generation technology in the industry. Stay tuned!
[11/2024] Rated as a Top Reviewer of NeurIPS 2024.
[07/2024] 1 paper is accepted by ECCV 2024.
[02/2024] 2 papers are accepted by CVPR 2024.
[07/2023] 1 paper is accepted by ICCV 2023.
[02/2023] 1 paper is accepted by CVPR 2023.
[05/2022] 1 paper is accepted by ICASSP 2022.

Publications

Below are my publications. (& means equal contribution, * refers to corresponding author.)
VBench-2.0: Advancing Video Generation Benchmark Suite for Intrinsic Faithfulness
Dian Zheng&, Ziqi Huang&, Hongbo Liu, Kai Zou, Yinan He, Fan Zhang, Yuanhan Zhang, Jingwen He, Wei-Shi Zheng*, Yu Qiao*, Ziwei Liu*,
ArXiv
[ArXiv] [Code] [Project Page]
(Video Generation Evaluation) Beyond Superficial Faithfulness, to target on next generation of video generation models, it's time to shift to evaluate the Intrinsic Faithfulness in the generated video. We comprehensively evaluate the latest video generation models in 18 dimensions. We hope it will be helpful to the development of video generation.
Panorama Generation From NFoV Image Done Right
Dian Zheng, Cheng Zhang, Xiao-Ming Wu, Cao Li*, Chengfei Lv, Jian-Fang Hu, Wei-Shi Zheng*,
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
[ArXiv] [Code] [Project Page]
(Panorama Generation) A novel decoupled paradigm to generate accurate distortion and appealing visual results at the same time. With our model, you can freely utilize real-world image with random NFoV, and even freely perform text-to-panorama generation. Enjoy it!
Diffusion Model for Volume based Stereo Matching
Dian Zheng, Xiao-Ming Wu, Zuhao Liu, Jingke Meng, Wei-Shi Zheng*
International Journal of Computer Vision (IJCV), 2025
[ArXiv] [Code]
(Stereo Matching) A lightweight, plug-and-play diffusion filter, boosting your stereo matching methods easily!
SpatialDreamer: Self-supervised Stereo Video Synthesis from Monocular Input
Zhen Lv, Yangqi Long, Congzhentao Huang, Cao Li, Chengfei Lv*, Hao Ren, Dian Zheng,
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
[ArXiv]
(Spatial Video Generation) A novel self-supervised stereo video generation paradigm that does not require paired stereo video data. I would call it the most advanced technology in the industry!
An Economic Framework for 6-DoF Grasp Detection
Xiao-Ming Wu&, Jia-Feng Cai&, Jian-Jian Jiang, Dian Zheng, Yi-Lin Wei, Wei-Shi Zheng*,
European Conference on Computer Vision (ECCV), 2024
[ArXiv] [Code]
(6-DoF Grasping) Speed up your 6-DoF grasping training 10x without performance drop!
Selective Hourglass Mapping for Universal Image Restoration Based on Diffusion Model
Dian Zheng, Xiao-Ming Wu, Shuzhou Yang, Jian Zhang, Jian-Fang Hu, Wei-Shi Zheng*,
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024
[ArXiv] [Code]
(Universal Image Restoration) Boosting the development of Universal Image Restoration by ajusting the diffusion algorithm!
Dexterous Grasp Transformer
Guo-Hao Xu&, Yi-Lin Wei&, Dian Zheng, Xiao-Ming Wu, Wei-Shi Zheng*,
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024
[ArXiv] [Code]
(Dexterous Grasping) The first discriminative framework for generating a diverse, high quality set of feasible dexterous grasp only in one pass!
Estimator Meets Equilibrium Perspective: A Rectified Straight Through Estimator for Binary Neural Networks Training
Xiao-Ming Wu, Dian Zheng, Zuhao Liu, Wei-Shi Zheng*
IEEE International Conference on Computer Vision (ICCV), 2023
[ArXiv] [Code]
(Binary Neural Networks) Proposing root-based STE for binary neural network training, achieving SOTA without bells and whistles!
Generating Anomalies for Video Anomaly Detection with Prompt-based Feature Mapping
Zuhao Liu, Xiao-Ming Wu, Dian Zheng, Kun-Yu Lin, Wei-Shi Zheng*
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023
[ArXiv]
(Video Anomaly Detection) The first syn2real anomaly feature generator.

Underwater Stereo Matching Via Unsupervised Appearance And Feature Adaptation Networks
Zhong Wei, Yazhi Yuan, Xinchen Ye*, Dian Zheng, Rui Xu
International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2022
[ArXiv]

Academic Service

Conference Reviewer: NeurIPS2025, ACMMM2025, ICCV2025, ICME2025, ICML2025, CVPR2025, ICLR2025, NeurIPS2024 (Top Reviewer), PRCV2023

Journal Reviewer: PR, CMV

Awards

First Prize, Academic Scholarship of Sun Yat-Sen University for Graduate Student (中山大学硕士研究生一等奖助金), 2022, 2023, 2024.

Xiaomi Grand Prize Scholarship of Sun Yat-Sen University (中山大学小米特等奖学金), 2024.


Last Update 12/25/2024. Thanks to Jon Barron.