Transfer Learning-based Road Damage Detection for Multiple Countries

TL;DR

This paper evaluates the transferability of Japanese road damage detection models to other countries, introduces a large heterogeneous dataset, and proposes generalized models for multi-country damage detection and classification.

Contribution

It assesses Japanese models' usability internationally, creates a large multi-country dataset, and develops generalized detection models for diverse road conditions.

Findings

Japanese models can be adapted to other countries

A large dataset of 26,620 images from multiple countries is introduced

Generalized models effectively detect and classify damages across countries

Abstract

Many municipalities and road authorities seek to implement automated evaluation of road damage. However, they often lack technology, know-how, and funds to afford state-of-the-art equipment for data collection and analysis of road damages. Although some countries, like Japan, have developed less expensive and readily available Smartphone-based methods for automatic road condition monitoring, other countries still struggle to find efficient solutions. This work makes the following contributions in this context. Firstly, it assesses the usability of the Japanese model for other countries. Secondly, it proposes a large-scale heterogeneous road damage dataset comprising 26620 images collected from multiple countries using smartphones. Thirdly, we propose generalized models capable of detecting and classifying road damages in more than one country. Lastly, we provide recommendations for readers, local agencies, and municipalities of other countries when one other country publishes its data and model for automatic road damage detection and classification. Our dataset is available at (https://github.com/sekilab/RoadDamageDetector/).