Billboard in Focus: Estimating Driver Gaze Duration from a Single Image

TL;DR

This paper presents an automated method to detect billboards and estimate driver gaze duration from single images, aiding in understanding distraction caused by roadside advertising.

Contribution

The study introduces a fully automated pipeline combining YOLO detection and DINOv2 features for estimating driver gaze duration without manual annotations or eye-tracking.

Findings

Achieved 94% mAP@50 in billboard detection

Estimated driver gaze duration with 68.1% accuracy on BillboardLamac

Validated results using Google Street View images

Abstract

Roadside billboards represent a central element of outdoor advertising, yet their presence may contribute to driver distraction and accident risk. This study introduces a fully automated pipeline for billboard detection and driver gaze duration estimation, aiming to evaluate billboard relevance without reliance on manual annotations or eye-tracking devices. Our pipeline operates in two stages: (1) a YOLO-based object detection model trained on Mapillary Vistas and fine-tuned on BillboardLamac images achieved 94% mAP@50 in the billboard detection task (2) a classifier based on the detected bounding box positions and DINOv2 features. The proposed pipeline enables estimation of billboard driver gaze duration from individual frames. We show that our method is able to achieve 68.1% accuracy on BillboardLamac when considering individual frames. These results are further validated using images collected from Google Street View.