A method for estimating roadway billboard salience

TL;DR

This work tackles the problem of estimating roadway billboard salience from a driver's viewpoint by combining billboard detection with saliency analysis. It compares YOLOv5 and Faster R-CNN for locating billboards, and evaluates UniSal and Spectral Residue saliency methods against eye-tracking data to assess significance. Findings show YOLOv5 generally outperforms Faster R-CNN after dataset-specific fine-tuning, while UniSal-based saliency aligns more closely with fixation data, yielding higher accuracy in billboard significance classification. The study contributes a reusable framework for driving-scene billboard analysis, supported by Mapillary Vistas and a custom eyetracking dataset, with potential implications for road-safety and advertising design.

Abstract

Roadside billboards and other forms of outdoor advertising play a crucial role in marketing initiatives; however, they can also distract drivers, potentially contributing to accidents. This study delves into the significance of roadside advertising in images captured from a driver's perspective. Firstly, it evaluates the effectiveness of neural networks in detecting advertising along roads, focusing on the YOLOv5 and Faster R-CNN models. Secondly, the study addresses the determination of billboard significance using methods for saliency extraction. The UniSal and SpectralResidual methods were employed to create saliency maps for each image. The study establishes a database of eye tracking sessions captured during city highway driving to assess the saliency models.

Figures (2)

• Figure 1: Image with annotated billboards from our dataset
• Figure 2: The scheme of the process of classification of the significance of billboards on the images