Food For Everyday Web Development
Jun 2020 - Present
- Develop web application that recommends cheap and suitable recipe with React.js;
- Collect 300+ ingredients’ price data daily for using Crawler and save to MongoDB.
- Configure Nginx to implement load balancing and reverse proxy to protect from DOS.
- Realize Asynchronous communication between users and recipe APIs by deploying Apache Kafka
- Optimize recommendation (nutrition, price, diet, preference) based on KKT conditions
- Utilized: Apache Kafka, Nginx, Digital Ocean, HTML/CSS, Optimization
System Platform Dev of Classification & Detection
Jan 2019-July 2019
- Developed deep Learning networks (CNN, ResNet) networks with PyTorch, achieved the goal of classification and detection for products with accuracy more than 95% within 0.003s
- Integrated visualized interface to label image and train networks based on PyQt5 and other Qt gadgets
- Combined labeling, training and application, simplified traditional deep learning theory-to-practice workflow, increasing overall efficiency by 67%
- Configured environment and test software through Docker in containers
- Utilized: pyinstaller, Docker, Git, CNN, R-CNN, ResNet
Turn-Based Strategy (TBS/SLG) game development (C, OpenCV)
Sep 2018-Dec 2018
- Developed the architecture the software; loaded data by decoding maneuver files.
- Revealed animation on console with OpenCV and tackled screen flashing by Double Buffering
- Configured observer pattern, defined dependencies among objects to update real time
- Utilized: C Programming, Data Structure Design, Observer Pattern, OpenCV