I am a PhD. student at Tsinghua University, Interdisciplinary Information Sciences, advised by Prof. Yang Gao. I obtained B.E. at Shanghai Jiao Tong University, majoring in automation(IEEE class). I love to study how the digital world percepts and controls the physical world. So, my research interests are robotics and embodied AI.
Favored programming language: Python, C++, Matlab. Like to program in an organized way: understandable variable names, clean comments, good file structure and git commit logs. Hate functions that written by other programmer but have no comments or docs.
🎖 Honors and Awards
- 2022.11 CUMCM, Provincial Second Prize
- 2022.10 Suzhou Yu Cai Scholarship
- 2022.10 SJTU Scholarship, Grade B
- 2022.4 COMAP, M Prize
- 2021.10 COSCO Scholarship, Third Award
- 2021.10 SJTU Scholarship, Grade B
📖 Educations
- 2024.09 - now, PhD. student, Tsinghua University, Beijing.
- 2020.09 - 2024.06, Undergraduate, Shanghai Jiao Tong University, Shanghai.
Score: 91.02/100 Rank: 1/28
💬 Language
- CET6 660
- TOEFL 104 (Reading 29, Listening 29, Speaking 22, Writing 24)
📝 Publications
CoPa: General Robotic Manipulation through Spatial Constraints of Parts with Foundation Models
Haoxu Huang, Fanqi Lin, Yingdong Hu, Shengjie Wang, Yang Gao
- CoPa is a novel framework that incorporates common sense knowledge embedded within foundation vision-language models (VLMs), such as GPT-4V, into the low-level robotic manipulation tasks.
- CoPa is capable of handling diverse open-set instructions and objects in a zero-training manner.
💻 Projects
Make Illustrator Relationship Visible
github
The final project for course ICE2604(电类工程导论). It is a cooperation project led by me. This project includes:
- Scraper We implemented a data scraper with Python and get information of illustrators and illustrations from pixiv
- Database The data are stored and organized in mysql.
- Analyze & Visualization We analyze the relationship among illustrators by tags attached to their works and followers. Then, we visualized these results as charts and a huge relation map.
- Website We designed an interactive website where users can look into statistics, search for illustrators and browse illustrations.
A P2P BBS
github
A BBS based on P2P network. Final project for Computer Network.
Fundus Images Processing
Final project for Digital Image Processing.
We first localize the macula lutea with CNN, the align and normalize the image. Next, the vessels are segmented with U-Net and are removed to get clean images with Fast Marching Method (FMM).