Hanshu YAN (严 汉书) (Pronouns: he/him/his)

About Me

Hi! I am a Research Scientist at ByteDance Singapore, where our team focuses on developing efficient and controllable video/image generation models and applying them to creative editing tasks. Previously, I completed my Ph.D. under the supervision of Prof. Vincent Y. F. Tan and Dr. Jiashi Feng at NUS. I also had the pleasure of collaborating with Dr. Jingfeng Zhang and Prof. Masashi Sugiyama from RIKEN-AIP. My research interests lie in machine learning and computer vision with the goal of exploring principled machine learning algorithms that can address challenging problems in engineering and science.

Academic Service

• Reviewer for ICML, ICLR, NeurIPS, CVPR, ACM Multimedia, etc.

Education

Ph.D,  Deep Learning	NUS	2022
M.Sc,  EE	NUS	2017
B.Eng, EE & Math	BUAA	2015

Selected Papers

( * Equally contributed; ^AIGC, ^ML robustness, ^Learning algorithms, ^Others )

• ^ MagicVideo-V2: Multi-Stage High-Aesthetic Video Generation. arXiv 2024
  Weimin Wang, Jiawei Liu, Zhijie Lin, Jiangqiao Yan, Shuo Chen, Chetwin Low, Tuyen Hoang, Jie Wu, Jun Hao Liew, Hanshu Yan, Daquan Zhou, Jiashi Feng


• ^ MagicAnimate: Temporally Consistent Human Image Animation using Diffusion Model . arXiv 2023
  Zhongcong Xu, Jianfeng Zhang, Jun Hao Liew, Hanshu Yan, Jia-Wei Liu, Chenxu Zhang, Jiashi Feng, Mike Zheng Shou


• ^ SAG: Towards Accurate Guided Diffusion Sampling through Symplectic Adjoint Method. arXiv 2023
  Jiachun Pan*, Hanshu Yan*, Jun Hao Liew, Jiashi Feng, Vincent Y. F. Tan


• ^ MagicAvatar: Multimodal Avatar Generation and Animation. arXiv 2023
  Jianfeng Zhang*, Hanshu Yan*, Zhongcong Xu*, Jiashi Feng, Jun Hao Liew*


• ^ MagicEdit: High-Fidelity and Temporally Coherent Video Editing. arXiv 2023
  Jun Hao Liew*, Hanshu Yan*, Jianfeng Zhang, Zhongcong Xu, Jiashi Feng


• ^ MagicProp: Diffusion-based Video Editing via Motion-aware Appearance Propagation. arXiv 2023
  Hanshu Yan*, Jun Hao Liew*, Long Mai, Shanchuan Lin, Jiashi Feng


• ^ AdjointDPM: Adjoint Sensitivity Method for Gradient Backpropagation of Diffusion Probabilistic Models. ICLR 2024
  Jiachun Pan*, Jun Hao Liew, Vincent Y. F. Tan, Jiashi Feng, Hanshu Yan*


• ^ MagicVideo: Efficient Video Generation With Latent Diffusion Models. arXiv 2023
  Daquan Zhou, Weimin Wang, Hanshu Yan, Weiwei Lv, Yizhe Zhu, Jiashi Feng


• ^ MagicMix: Semantic Mixing with Diffusion Models. arXiv 2022
  Jun Hao Liew*, Hanshu Yan*, Daquan Zhou, Jiashi Feng


• ^ Learning a robust foundation model against clean-label data poisoning attacks at downstream tasks. Neural Networks 2023
  Ting Zhou, Hanshu Yan*, Bo Han, Lei Liu, Jingfeng Zhang


• ^ Pre-training Robust Feature Extractor Against Clean-label Data Poisoning Attacks. NeurIPS Workshop on Machine Learning Safety 2022
  Ting Zhou, Hanshu Yan*, Lei Liu, Jingfeng Zhang, Bo Han


• ^ Towards Adversarially Robust Deep Image Denoising. IJCAI 2022
  Hanshu Yan, Jingfeng Zhang, Jiashi Feng, Masashi Sugiyama, Vincent Y. F. Tan


• ^ CIFS: Improving Adversarial Robustness of CNNs via Channel-wise Importance-based Feature Selection. ICML 2021  [Code]  [Poster]
  Hanshu Yan, Jingfeng Zhang, Gang Niu, Jiashi Feng, Vincent Y. F. Tan, Masashi Sugiyama


• ^ RAIN: Robust and Accurate Classification Networks with Randomization and Enhancement. arXiv 2020  [Code]
  Jiawei Du, Hanshu Yan, Joey Tianyi Zhou, Rick Siow Mong Goh, Jiashi Feng


• ^ On Robustness of Neural Ordinary Differential Equations. ICLR 2020 Spotlight  [Code]
  Hanshu Yan, Jiawei Du, Vincent Y. F. Tan, Jiashi Feng


• ^ Towards Better Time Series Contrastive Learning: A Dynamic Bad Pair Mining Approach. ICLR 2024
  Xiang Lan, Hanshu Yan, Shenda Hong, Mengling Feng


• ^ Efficient Sharpness-aware Minimization for Improved Training of Neural Networks . ICLR 2022  [Code]
  Jiawei Du, Hanshu Yan, Jiashi Feng, Joey Tianyi Zhou, Liangli Zhen, Rick Siow Mong Goh, Vincent Y. F. Tan


• ^ Towards Understanding Why Lookahead Generalizes Better Than SGD and Beyond. NeurIPS 2021
  Pan Zhou, Hanshu Yan, Xiaotong Yuan, Jiashi Feng, Shuicheng Yan


• ^ Information-Theoretic Characterization of the Generalization Error for Iterative Semi- Supervised Learning. Journal of Machine Learning Research, 2022.
  Haiyun He, Hanshu Yan, Vincent Y. F. Tan


• ^ Deep learning-guided fiberoptic Raman spectroscopy enables real-time in vivo diagnosis and assessment of nasopharyngeal carcinoma. Analytical Chemistry 2021
  Chi Shu, Hanshu Yan, Wei Zheng, Kan Lin, Anne James, Selvarajan, Sathiyamoorth, Chwee Ming Lim, Zhiwei Huang


• ^ Recovering the Unbiased Scene Graphs from the Biased Ones. ACM MM 2021  [Code]
  Meng-Jiun Chiou, Henghui Ding, Hanshu Yan, Changhu Wang, Roger Zimmermann, Jiashi Feng


• ^ Modeling Trajectories with Neural Ordinary Differential Equations. IJCAI 2021
  Yuxuan Liang, Kun Ouyang, Hanshu Yan, Yiwei Wang, Zekun Tong and Roger Zimmermann


• ^ Adversarial Domain Adaptation with Prototype-Based Normalized Output Conditioner. IEEE Transactions on Image Processing 2021
  Dapeng Hu, Jian Liang, Qibin Hou, Hanshu Yan, Yunpeng Chen

Experiences

Research Scientist	ByteDance, Singapore	07/2022 - Present
Research Intern	Sea AI Lab, Singapore	03/2021 - 09/2021
Teaching Assistant	NUS, Singapore	08/2018 - 12/2021
Reserach Engineer	NUS, Singapore	07/2017 - 07/2018

Scholarships & Awards

• NUS Research Scholarship, 2018-2022
• Outstanding Graduating Student (B.Eng @ BUAA), 2015

Coursework

@ School

• EE5907 Pattern Recognition; EE4603 Biomedical Imaging Systems
• EE5904R Neural Networks; EE6934 Deep Learning
• EE5137 Stochastic Processes; EE5139 Information Theory
• EE5101R Linear Systems; EE5103R Computer Control Systems
• EE5702R Advanced Power System Analysis

@ Home

• NUS, MA3209 Metric space
• CMU, 36-705 Intermediate Statistics
• Stanford, CS224n Natural Language Processing with Deep Learning

[tmp!]