Curriculum
Computer Science for Societal Challenges and Innovation, XL series
Grant sponsor
China Scholarship Council, CSC Scholarship
Supervisor
Lamberto Ballan
Co-supervisor
to be defined
Contact
jing.mi@studenti.unipd.it

Project description
Multi-agent behavior prediction in the surrounding environment has attracted increasing attention in fields such as autonomous driving, social robotics, computer vision, human-robot cooperation, intelligent surveillance, and robot navigation systems, etc. Behaviour prediction covers a wide range of directions. My research focuses on trajectory prediction and traffic behaviour prediction. Multi-agent trajectory prediction is one of the most mainstream research directions. The major research contents in multi-agent behaviour prediction include modelling the interactions between multi-agents, and the surrounding environment influence, designing prediction frameworks, and optimizing encoder-decoder and multimodal prediction methods. The major challenges lie in complex heterogeneous interactions between multi-agents, multimodal behaviours in the future, improving efficiency and interpretability of the prediction model, predicting socially acceptable and physically acceptable behaviours, and long-term predictions.