Xiaohan Zou 邹笑寒
Xiaohan’s current research interests lie primarily in understanding and improving continuity, generalizability, efficiency and other key properties of deep learning. She is also interested in vision-language and video models.
Previously, she received her master’s degree in Computer Science at Boston University and bachelor’s degree in Software Engineering at Tongji University. She was fortunate to work with Prof. Bryan Plummer (BU) on continual learning and parameter sharing, Prof. Tong Lin (PKU) on continual learning and machine translation, and Prof. Qinpei Zhao (Tongji) on image aesthetic captioning.
Check her Curriculum Vitae for more details.
Education
| M.S. in Computer Science, Boston University | 2021-2023 |
| B.Eng. in Software Engineering, Tongji University Thesis: Food Image Aesthetic Assessment and Captioning | 2016-2020 |
Publications
TokenFlow: Rethinking Fine-grained Cross-modal Alignment in Vision-Language Retrieval
Xiaohan Zou, Changqiao Wu, Lele Cheng, and Zhongyuan Wang
Preprint, 2022
paper / abstract
Efficient Meta-Learning for Continual Learning with Taylor Expansion Approximation
Xiaohan Zou, and Tong Lin
International Joint Conference on Neural Networks (IJCNN), 2022
Oral Presentation
To be an Artist: Automatic Generation on Food Image Aesthetic Captioning
Xiaohan Zou, Cheng Lin, Yinjia Zhang, and Qinpei Zhao
International Conference on Tools with Artificial Intelligence (ICTAI), 2020
Oral Presentation
A Survey on Application of Knowledge Graph
Xiaohan Zou
International Conference on Control Engineering and Artificial Intelligence (CCEAI), 2020
paper / abstract
Experience
| Machine Learning Engineer Intern, Kuaishou, Beijing, China | 2021–2022 |
| Research Intern, Peking University, Beijing, China (with Prof. Tong Lin) | 2020–2022 |
| Software Engineer Intern, China Electronics Technology Group Corporation | 2020–2021 |
| Game Engineer Intern, Banana Interactive, Shanghai, China | 2019–2020 |
| Research Intern, Peking University, Beijing, China (with Prof. Tong Lin) | 2018 |
Talks
- Meta / Few-shot Learning, Kuaishou, 08/2021
- Continual Learning: Meta Continual Learning & Task Free Settings, Peking University, 08/2020
Miscellaneous
🚀 This personal website is built on Vitesse, with the power of Vue 3, Vite and UnoCSS
🧐 Here are two “interesting” versions of my portfolio/cv: portfolio.zxh.io, resume.zxh.io
🎃 Renovamen is a Latin word means renewal
🖥 Ex-OIer/ACMer
🛠 Like to contribute to open-source machine learning and frontend projects in spare time
🥎 Used to be a member of the softball team of Tongji University
🌭 My dream:
while(sleeping){money++;}🕹️ Currently interested in constructing buildings in Minecraft
📜 大概率更新不及时的中文简历