Responsibilities include:

• Trained, optimized and deployed trajectory prediction models with Waymo Open Motion Dataset and private-owned datasets on Nvidia Orin
• Developed basic software framework for prediction module based on Baidu Apollo open-source framework (ROS-based) to serve as a foundation for further algorithm development
• Built a data extraction pipe to prepare data for training and testing
• Provided mentorship to 2 interns and set step-by-step goals to help them get familiar with algorithm optimization based on case feedback

Responsibilities include:

• Proposed a new trajectory prediction algorithm by incorporating dynamic models to optimize the current rule-guided algorithm to smooth the vehicle trajectory
• Conducted thorough research and literature review on interaction-aware trajectory prediction models
• Applied reinforcement learning algorithms in planning decisions such as the pull-over spot selection

Responsibilities include:

• Test Item Acquisition Algorithm for Semiconductor
• Reinforcement Learning for Vehicle Routing Problems
• Multi-modal Trajectory Prediction

Responsibilities include:

• GNN for network performance KPIs prediction