• Content-recognition algorithm optimization:
• Client application profiling, analysis and algorithm improving to reach CPU usage reduction
• Modified C++ components, wrote unit-tests with Python and evaluated performance change
• Integrated two of company’s task management tools (Asana and FogBugz) through API scripting
• Built a tool set based on C++ and Python to realize batched test dataset creation and annotation
• Worked on hardware product maintenance and QA
• Developed a web application for video files metadata editing, including the following components
• An HTML5 based video player with sequenced video frames preview window for accurate video frame locating
• Implemented metadata editing and file management user interfaces
• Built the backend controller based on Pyramid framework and MongoDB
• Scripted for simulation and analysis of content-recognition algorithms, verified improvement on algorithm performance with Python