Biography
I am a Machine Learning Engineer at TikTok, specializing in multimodal Large Language Model (LLM) applications.
I earned my Ph.D. in Computer Science from Duke University in December 2023, where I focused on video motion tracking and stereo vision under the supervision of Professor Carlo Tomasi.
Prior to that, I graduated with a B.S. in Computer Science from Nanjing University in June 2018, where I conducted undergraduate research under Professor Zhi-Hua Zhou.
Work Experiences
- Machine Learning Engineer, TikTok Trust & Safety Algo team. San Jose CA, Feb 2024 - Present.
- Research Scientist Intern, Meta Reality Labs Core AI team. Seattle WA, May 2023 - Dec 2023.
- Research Engineer Intern, Meta Reality Labs Surreal team. Redmond WA, May 2022 - Aug 2022.
- Research Scientist Intern, Meta Reality Labs Agios team. Redmond WA, May 2020 - Aug 2020.
Publications
- Shuai Yuan, Lei Luo, Zhuo Hui, Can Pu, Xiaoyu Xiang, Rakesh Ranjan, and Denis Demandolx. UnSAMFlow: Unsupervised Optical Flow Guided by Segment Anything Model. The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024.
[code] |
[appendix] |
[poster] |
[video] |
[cite]
- Shuai Yuan, Shuzhi Yu, Hannah Kim, and Carlo Tomasi. SemARFlow: Injecting Semantics into Unsupervised Optical Flow Estimation for Autonomous Driving. International Conference on Computer Vision (ICCV), 2023.
[code] |
[appendix] |
[poster] |
[video] |
[cite]
- Shuai Yuan, Xian Sun, Hannah Kim, Shuzhi Yu, and Carlo Tomasi. Optical Flow Training under Limited Label Budget via Active Learning. European Conference on Computer Vision (ECCV), 2022.
[code] |
[appendix] |
[poster] |
[video] |
[cite]
- Shuai Yuan and Carlo Tomasi. UFD-PRiME: Unsupervised Joint Learning of Optical Flow and Stereo Depth through Pixel-Level Rigid Motion Estimation. arXiv Preprint, 2023.
[appendix] |
[cite]
- Yimu Wang*, Xuye Liu*, Wei Pang*, Li Ma*, Shuai Yuan*, Paul Debevec, and Ning Yu. Survey of Video Diffusion Models: Foundations, Implementations, and Applications. arXiv Preprint, 2025.
(* Equal contribution)
- Fanjie Kong, Shuai Yuan, Weituo Hao, and Ricardo Henao. Mitigating Test-Time Bias for Fair Image Retrieval. Annual Conference on Neural Information Processing Systems (NeurIPS), 2023.
[code] |
[appendix] |
[poster] |
[video] |
[cite]
- Xiang Wang, Shuai Yuan, Chenwei Wu, and Rong Ge. Guarantees for Tuning the Step Size using a Learning-to-Learn Approach. International Conference on Machine Learning (ICML), 2021.
[code] |
[appendix] |
[poster] |
[video] |
[cite]
- Yimu Wang, Shuai Yuan, Bo Xue, Xiangru Jian, Wei Pang, Mushi Wang, and Ning Yu. DREAM: Improving Video-Text Retrieval Through Relevance-Based Augmentation Using Large Foundation Models. The Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL), 2025.
[cite]
- Kelsey Lieberman, Shuai Yuan, Swarna Kamlam Ravindran, and Carlo Tomasi. Optimizing for ROC Curves on Class-Imbalanced Data by Training over a Family of Loss Functions. arXiv Preprint, 2024.
[appendix] |
[cite]
- Shuzhi Yu, Hannah Kim, Shuai Yuan, and Carlo Tomasi. Unsupervised Flow Refinement near Motion Boundaries. British Machine Vision Conference (BMVC), 2022.
[code] |
[appendix] |
[poster] |
[video] |
[cite]
- Hannah Kim, Shuzhi Yu, Shuai Yuan, and Carlo Tomasi. Cross-Attention Transformer for Video Interpolation. Asian Conference on Computer Vision (ACCV) - Workshop on Vision Transformers: Theory and Application, 2022. Best Paper Award.
[code] |
[video] |
[cite]
Conference Reviewer
- Computer Vision: CVPR 2024, 2025; ICCV 2025; ECCV 2024; ACCV 2022.
- Machine Learning: ICML 2024, 2025; ICLR 2025; NeurIPS 2023.
- Artificial Intelligence: AAAI 2025.
Teaching Assistant
Personal Interests
- Sports/Workouts: gym training, swimming, hiking, badminton, table tennis, tennis, volleyball, etc.
- Handcrafting: lego-style blocks, mechanical, electronic, metal, crochet, etc.
- Drama/Musicals: Les Misérables, Dear Evan Hansen, Chinese original musicals, etc.