It’s important that self-driving cars quickly detect other cars or pedestrians sharing the road. Researchers at Carnegie Mellon University have shown that they can significantly improve detection accuracy by helping the vehicle also recognize what it doesn’t see that is the empty space.
Peiyun Hu, a Ph.D. student’s work enables a self-driving car’s perception systems to consider visibility as it reasons about what its sensors are seeing.
Reasoning about visibility is already used when companies build digital maps. “Map-building fundamentally reasons about what’s empty space and what’s occupied,” said Deva Ramanan, an associate professor of robotics and director of the CMU Argo AI Center for Autonomous Vehicle Research. He further adds that the same concept does not occur for live, on-the-fly processing of moving obstacles that are at traffic speeds.
Hu and his colleagues borrow techniques from map-making to help the system reason about visibility when trying to recognize objects. However, their technique outperforms the current highest performing system. Hu’s system improves detection by 10.7% for cars, 5.3% for pedestrians, 7.4% for trucks, 18.4% for buses and 16.7% for trailers.
The research is currently being presented at the Computer Vision and Pattern Recognition (CVPR) virtual conference, June 13–19.
[Read More] Source: Carnegie Mellon University, School of Computer Science