Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PaveSync: A Unified and Comprehensive Dataset for Pavement Distress Analysis and Classification

Blessing Agyei Kyem, Joshua Kofi Asamoah, Anthony Dontoh, Andrews Danyo, Eugene Denteh, Armstrong Aboah

TL;DR

PaveSync tackles the fragmentation of pavement-distress datasets by introducing a large-scale, multi-perspective benchmark with standardized annotations, spanning 13 distress types across seven countries. The dataset aggregates diverse sources, validates annotation quality via stratified checks, and provides 52,747 images with 135,277 bounding boxes, split 90/10 for training and evaluation. It benchmarks seven state-of-the-art detectors (YOLOv8–v12, Faster R-CNN, DETR) to reveal technology-specific strengths and guide fair cross-model comparisons, including zero-shot transfer scenarios. Publicly released with preprocessed annotations and documentation, PaveSync aims to advance generalizable pavement monitoring, enable domain adaptation, and support safer, more cost-effective road maintenance through robust, reproducible research.

Abstract

Automated pavement defect detection often struggles to generalize across diverse real-world conditions due to the lack of standardized datasets. Existing datasets differ in annotation styles, distress type definitions, and formats, limiting their integration for unified training. To address this gap, we introduce a comprehensive benchmark dataset that consolidates multiple publicly available sources into a standardized collection of 52747 images from seven countries, with 135277 bounding box annotations covering 13 distinct distress types. The dataset captures broad real-world variation in image quality, resolution, viewing angles, and weather conditions, offering a unique resource for consistent training and evaluation. Its effectiveness was demonstrated through benchmarking with state-of-the-art object detection models including YOLOv8-YOLOv12, Faster R-CNN, and DETR, which achieved competitive performance across diverse scenarios. By standardizing class definitions and annotation formats, this dataset provides the first globally representative benchmark for pavement defect detection and enables fair comparison of models, including zero-shot transfer to new environments.

PaveSync: A Unified and Comprehensive Dataset for Pavement Distress Analysis and Classification

TL;DR

PaveSync tackles the fragmentation of pavement-distress datasets by introducing a large-scale, multi-perspective benchmark with standardized annotations, spanning 13 distress types across seven countries. The dataset aggregates diverse sources, validates annotation quality via stratified checks, and provides 52,747 images with 135,277 bounding boxes, split 90/10 for training and evaluation. It benchmarks seven state-of-the-art detectors (YOLOv8–v12, Faster R-CNN, DETR) to reveal technology-specific strengths and guide fair cross-model comparisons, including zero-shot transfer scenarios. Publicly released with preprocessed annotations and documentation, PaveSync aims to advance generalizable pavement monitoring, enable domain adaptation, and support safer, more cost-effective road maintenance through robust, reproducible research.

Abstract

Automated pavement defect detection often struggles to generalize across diverse real-world conditions due to the lack of standardized datasets. Existing datasets differ in annotation styles, distress type definitions, and formats, limiting their integration for unified training. To address this gap, we introduce a comprehensive benchmark dataset that consolidates multiple publicly available sources into a standardized collection of 52747 images from seven countries, with 135277 bounding box annotations covering 13 distinct distress types. The dataset captures broad real-world variation in image quality, resolution, viewing angles, and weather conditions, offering a unique resource for consistent training and evaluation. Its effectiveness was demonstrated through benchmarking with state-of-the-art object detection models including YOLOv8-YOLOv12, Faster R-CNN, and DETR, which achieved competitive performance across diverse scenarios. By standardizing class definitions and annotation formats, this dataset provides the first globally representative benchmark for pavement defect detection and enables fair comparison of models, including zero-shot transfer to new environments.
Paper Structure (24 sections, 4 equations, 3 figures, 3 tables)

This paper contains 24 sections, 4 equations, 3 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Studies
  3. Dataset
  4. Data Sources and Composition
  5. Data Standardization
  6. Annotation Validation
  7. Final Dataset Composition
  8. Methodology
  9. Data Preprocessing and Augmentation
  10. Experimental Setup
  11. Deep Learning Models for Benchmarking
  12. YOLO Variants
  13. Faster R-CNN
  14. DETR: Detection Transformers
  15. Evaluation Metrics
  16. ...and 9 more sections

Figures (3)

  • Figure 1: Distribution of the PaveSync dataset across different countries
  • Figure 2: Different imaging orientations in the dataset, including ground-level, aerial, and top-down views
  • Figure 3: Sample images from the dataset captured under different weather conditions