Evaluating the Significance of Outdoor Advertising from Driver's Perspective Using Computer Vision

TL;DR

This paper addresses estimating the significance of roadside billboards from a driver's perspective using computer vision and eyetracking data. It introduces the BillboardLamac dataset and a full pipeline for billboard detection, tracking, saliency estimation, and billboard-significance classification. The strongest results are reported for tracking with a $HOTA$ score of $38.5$, and a Random Forest classifier achieves a test accuracy of $75.8\%$ in classifying billboards into three gaze-duration-based categories, with duration, saliency features, and billboard size being the most influential. The approach enables eyetracking-free deployment and offers actionable guidance for billboard design and placement to balance advertising impact with road safety, while highlighting directions for dataset expansion and feature augmentation.

Abstract

Outdoor advertising, such as roadside billboards, plays a significant role in marketing campaigns but can also be a distraction for drivers, potentially leading to accidents. In this study, we propose a pipeline for evaluating the significance of roadside billboards in videos captured from a driver's perspective. We have collected and annotated a new BillboardLamac dataset, comprising eight videos captured by drivers driving through a predefined path wearing eye-tracking devices. The dataset includes annotations of billboards, including 154 unique IDs and 155 thousand bounding boxes, as well as eye fixation data. We evaluate various object tracking methods in combination with a YOLOv8 detector to identify billboard advertisements with the best approach achieving 38.5 HOTA on BillboardLamac. Additionally, we train a random forest classifier to classify billboards into three classes based on the length of driver fixations achieving 75.8% test accuracy. An analysis of the trained classifier reveals that the duration of billboard visibility, its saliency, and size are the most influential features when assessing billboard significance.

Figures (7)

• Figure 1: A frame from the dataset BillboardLamac with annotated billboards (ID 67 and 68) and their corresponding bounding boxes.
• Figure 2: Map of the round trip taken by the participants of the user study displayed in Google Maps.
• Figure 3: Fixations captured by Tobii Pro Glasses displayed on a frame from a video captured in the user study.
• Figure 4: Frame from our BillboardLamac dataset with 3 billboards detected by the Yolov8large model RangeKing23 trained on a reduced version of the Mapillary Vistas dataset Neuhold17.
• Figure 5: Tracking by detection. Objects are first detected via bounding boxes which are then associated into separate tracks.