Smart car algorithm will detect pedestrians in real-time

(Image Credit: iStockPhoto/olaser)

Engineers at the University of California, San Diego have developed a pedestrian detection system for smart vehicles which can detect pedestrians in near real-time to help ensure the streets remain safe in the era of semi-autonomous and driverless cars.

Results we’re obtaining with this new algorithm are substantially better for real-time, accurate pedestrian detection 

The system makes use of deep-learning models and "computer vision" systems to help the vehicle to better understand what's happening around them – not just other vehicles. It combines a traditional computer vision classification architecture – known as cascade detection – with deep learning models to detect pedestrians with a high-degree of accuracy. 

Similar technologies in-use today will break down an image into small windows that are processed by a classifier that signals the presence or absence of a pedestrian. This method is found to be inefficient, as pedestrians appear in different sizes depending on distance to the camera. 

In contrast, this system can reliably detect pedestrians at a rate of 2 to 4 frames per second – or roughly as well as the human eye can – due to a smart algorithm which can filter out areas where human activity is not detected such as the sky or surrounding buildings. By filtering out the unnecessary information, the system can keep an eye on imminent threats.

The innovative technology was developed by Nuno Vasconcelos, Electrical Engineering Professor, and his team at the UC San Diego Jacobs School of Engineering. “No previous algorithms have been capable of optimizing the trade-off between detection accuracy and speed for cascades with stages of such different complexities. In fact, these are the first cascades to include stages of deep learning. The results we’re obtaining with this new algorithm are substantially better for real-time, accurate pedestrian detection,” said Vasconcelos. 

His work, titled “Learning Complexity-Aware Cascades for Deep Pedestrian Detection,” was presented Dec. 15, 2015 at the International Conference on Computer Vision in Santiago, Chile. The research team also included Zhaowei Cai of UC San Diego and Mohammed Saberian of Yahoo Labs. 

Do you think we can rely on algorithms to ensure pedestrian safety? Let us know in the comments.

Leave a comment


This will only be used to quickly provide signup information and will not allow us to post to your account or appear on your timeline.