Automation and AI-integration in an injection mold manufacturing pipeline.

Worked on developing different features of an AI background music development platform. Key contributions included: • Researched the landscape of background music generation tools, evaluating various models for video analysis and music generation to propose a Minimum Viable Product (MVP). • Developed features for uploading videos to the platform and generating text-based descriptions. • Implemented functionality to extract videos from YouTube URLs and detect video intros and outros. • Assessed the feasibility of different multi-modal AI models in terms of accuracy, efficiency, and resource requirements.

更多

Solved various practical problems while building a mobile autonomous warehouse robot for a dynamic environment. Key contributions included: • Fused different sensor data (LiDAR, IMU, and GPU) for improved localization. • Developed a PID controller node for precise docking at the charging station. • Refined the Adaptive Pure Pursuit algorithm for obstacle avoidance.

更多

Worked on a project conducted jointly by the lab and Tomocube, which is about generating 3D stained image from 3D RI (refractive index) data acquired from cell samples supplied by hospitals around Korea. A data driven approach to train 3D convolutional neural networks to translate RI tomograms to FL (fluorescence) tomograms was used. Key contributions included: • data preprocessing • image labelling • building the ML model • training

更多