[email protected] · https://github.com/densechen · https://densechen.github.io/

I am Dengsheng Chen, an experienced AI researcher currently working at Meituan. I have previously interned at Tencent AI Lab, ByteDance, and Ecovacs Nanjing AI Research Institute. My expertise lies in a range of cutting-edge technologies including AIGC, 3DV, NeRF, LLM, and FedML. I have also authored several research papers published in top-tier conferences such as CVPR, ICLR, and AAAI.

I am now looking forward to joining a doctoral program that values technical excellence and innovation, where I can further my research in interactive AI.

• Master's Degree in Computer Science, National University of Defense Technology, 2019-2021
• Bachelor's Degree in Computer Science, Fuzhou University, 2015-2019

• Meituan, AI Researcher, 2022-Present
• ByteDance (TikTok), AI Intern, 2021
• Ecovacs Nanjing AI Research Institute, AI Intern, 2019
• Tencent AI Lab, AI Intern, 2018

Over the years, I have contributed to a number of research papers in the field of AI, with several currently under review for conferences like CVPR and ICLR. My research interests and contributions span across various topics including animation, video sampling, 3D shape manipulation, optimization, and more. Here are a few of my notable research works:

• "Animating General Image with Large Visual Motion Model". Under review.
• "Zero-Shot Video Sampling from Image". Under review.
• "Geometric Imitation Models for 3D Shape Manipulation". Under review.
• "Real3D: The Curious Case of Neural Scene Degeneration". Proceedings of the 38th Annual AAAI Conference on Artificial Intelligence. 2024.
• "Elastic Aggregation for Federated Optimization". Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023. code
• "Rethinking skip connection model as a learnable Markov chain". Proceedings of the International Conference on Learning Representations. 2023. code
• "Learning canonical shape space for category-level 6d object pose and size estimation." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020. code

For more details, please refer to my Google Scholar profile.

Beyond my research, I am also a core developer of the OpenFed framework and the OpenMM series of tools. I have contributed code to PyTorch3D and have participated in the peer review process for several journals and conferences.

If you wish to connect or have any queries, feel free to reach me via email at [email protected] or visit my personal website for more information.