Address: Room 809, Building of Computer Science and Technology, Nanjing University
Phone: (+86) 18805156508
- B.E. in Software Engineering
Software Institute, Nanjing University
GPA: 4.54⁄5.0, Rank: 2⁄212
Nanjing University Undergraduate Innovation Program, 2016 - 2017
Advisor: Prof. Jia Bei
I was trained to read CV and AI papers, focusing on monocular depth estimation.
Intellisense node: a nodejs API recommendation plugin for vscode. A programmer can use our tool to find a recommended API to finish his task. Our tool has 3 inputs: a task description in natural language, expected input/output type of the API, and the expected position to insert the code. Our backend is a database server which stores the API documentation as well as API usages (which we mined from Stack Overflow) of 3 libraries. We compare the surrounding control structures of the targeted position with those of online API usage code to leverage contextual information.
I intended to build the tool as an implementation exercise of my API research reading, trying to bring research ideas into the wild. Type and context were already used to implement auto-completion, and a whole line of work recognized code examples on Stack Overflow as complementary documentation (e.g.)
After I and my friends finished this project in July 2018, Microsoft officially released its IntelliCode in August, which is also a recommendation tool based on type, context, and a mined repository of practice. We were thrilled to hear that.
Saber car: a Raspeberry Pi autonomous car. During my Embedded System course, I and my roommate participated in a car racing competition, running our program on a prebuilt car (the motoring API was exposed as a C library to us), using our own webcam. I leveraged my CV knowledge to design a robust processing pipeline for our car. We got the 3rd place among 30 teams.
Supic: an image processing Electron application backed up by two deep learning models. (Depth estimation and Super-resolution) I and my teammates developed it based on latest DL papers in an Agile training course at the end of our Innovation Program. At the course, our project received the highest score.
Colony: a highly competitive libGDX game which I almost wrote on my own in my freshman year. I and my friends crafted this for a freshman game competition, and won the 2nd prize among 50 teams.
Open Source Experience
- Aside from disclosing my own projects, I fixed a bug in one CVPR 2017 paper, and a bug in Microsoft vscode
- Tung OOCL Scholarship, The Tung Foundation & Orient Overseas Container Line Ltd., 2018
- National Scholarship, Ministry of Education of China, 2017
- Huawei Scholarship, Huawei Technologies, 2016