CyLab Student Seminar

  • Robert Mehrabian Collaborative Innovation Center
  • 2101
  • Ph.D. Student
  • Department of Electrical Engineering and Computer Sciences
  • Carnegie Mellon University

A Collaborative Relative and Absolute Localization (CORAL) Framework for Self-Driving Vehicles

A self-driving vehicle needs localization to estimate the runtime position, orientation, and velocity to solve two problems: What is the vehicle's location? What does the vehicle need to do next, given its current location? This localization function must support two distinct requirements. One is path planning to avoid obstacles in the immediate local environment of the self-driving vehicle. This functionality works in a local coordinate system and needs relative localization. The second requirement is for navigation, which works in a global coordinate system and requires absolute localization. We propose collaboration between relative and absolute localization to bound the accumulated error from relative localization and lower the adverse effects of measurement interference on absolute localization. Thus, localization accuracy and robustness can be improved in both the local and global coordinate systems.

Mengwen He is a Ph.D. student at Carnegie Mellon University, Electrical and Computer Engineering Department. His research interests focus on the perception of self-driving vehicle, including vehicle localization, object detection and tracking, multi-sensor calibration, and scene understanding. He received his B.A. and M.S. in Schools of Electronics Engineering and Computer Science, Peking University. Before attending CMU, he was a research assistant at Nagoya University in Japan, and he was involved in the mobility research of the Center of Innovation (COI) program.

This is a practice talk for Mengwen's ECE qualifying exam presentation.

For More Information, Please Contact: