Joint Perception and Motion Prediction System for Autonomous Vehicles

The project is concerned with delivering a real-world joint perception prediction and tracking system that outperforms the current system implemented in the Uber ATG software stack. This will be achieved by implementing a production version of a state-of-the-art academic paper developed by the research team, testing it against real world scenarios, and then modifying the neural network, inputs, architecture, and operations to obtain better within system results.

Intern: 
Zhen Gou
Faculty Supervisor: 
Murat Erdogdu
Province: 
Ontario
Partner University: 
Discipline: 
Program: