Road Surface Defect

Road surface defect detection aims to automate the identification and classification of pavement damage using various data sources, primarily images but increasingly incorporating LiDAR and other sensor data. Current research focuses on improving the accuracy and generalizability of deep learning models, including convolutional neural networks and vision language models, often employing techniques like attention mechanisms and domain generalization to handle diverse datasets and varying conditions. This research is crucial for enhancing road safety, optimizing maintenance schedules, and reducing economic burdens associated with road damage, particularly in resource-constrained settings.