Friday 07 March 2025
Researchers have developed a new method for determining the significance of billboards, which could help reduce distractions on the road and improve driver safety. The approach uses computer vision techniques to detect billboards in images and then assess their saliency – or attention-grabbing power.
To develop this method, scientists fine-tuned two object detection models, YOLOv5 and Faster R-CNN, using a dataset of images taken from the dashboard of a car. They found that YOLOv5 outperformed Faster R-CNN in detecting billboards, particularly small ones.
Next, the researchers used two saliency extraction methods – UniSal and SpectralResidual – to generate maps showing where drivers were likely to look at each billboard. These maps were then compared to fixation data collected using an eyetracker device.
The results showed that both methods performed well in detecting areas of high attention, with UniSal producing more accurate results. The study’s findings suggest that by analyzing the saliency of billboards, it may be possible to identify those that are most likely to distract drivers and take action to reduce their visibility or impact on road safety.
The method could also have applications beyond road safety, such as in advertising and marketing research. By understanding what draws our attention, businesses may be able to design more effective campaigns that engage with consumers without causing distractions.
The study’s authors hope that their work will contribute to the development of more effective methods for reducing driver distractions and improving road safety. By combining computer vision techniques with eyetracking data, they have taken an important step towards achieving this goal.
Cite this article: “Measuring Billboard Saliency to Improve Driver Safety and Reduce Distractions”, The Science Archive, 2025.
Billboards, Computer Vision, Driver Safety, Object Detection, Saliency, Attention-Grabbing Power, Road Safety, Advertising, Marketing Research, Distraction Reduction
