Hi there!I'm Tianyu ๐ I'm currently doing algorithm testing at Aibee ๐ to ensure and improve the quality of image algorithms. Recently, I'm interested in Full-Stack technology, UI design, 3D arts, and games of course.
Test skills
- Basic: Python, Shell, Docker
- Automation: Airflow, GitLab-CI
- Analysis & Visualization: Numpy, Pandas, Opencv, Plotly
Full-Stack
- Basic: HTML, CSS, JavaScript
- Back-End: Express, MongoDB
- Front-End: Vue3, Nuxt
Others
- Language: Chinese(Mother tongue), English(IELTS: 7)
- 2D Design: Figma
- 3D Design: Blender, CATIA
- Responsible for 5+ models' testing and giving test reports including badcase analysis and data visualization
- Face recognition model
- Face anti-spoofing model
- Face mask classification model
- Body 21-attributes model
- Car brand classification model
- Liases with algorithm researchers and improve the performance of models
- Design, write and debugged dev-tools to automate model test procedures, and boost the model verification flow
- Leading a four people group, developing a data management system called Data Manager, which helps testers managing millions of images' data and hundreds of benchmarks. Increasing dataset management efficiency by 20%.
- Cooperating with algorithms researchers, pass the key algorithm certification: BCTC face anti-spoofingcertification.
- Optimized the face recognition test tools with Faiss, shortening test time nearly 4X.
BSc. Vehicle Engineering 2015.9-2019.6 @South China University of Technology , China
Designed & Coded by Tianyu @2022