LDA-YOLO11: a lightweight detail-attention detector for edge-deployable road surface defect detection
- 1Department of Computer Science, Dongshin University
- 2Department of Automotive Engineering, Yichang Vocational College of Science and Technology
- 3Department of Electronic and Information Engineering, Anhui University
Abstract
Road surface defects, such as cracks, pavement damage and potholes, shorten road service life and affect driving safety and preventive maintenance. Existing lightweight object detectors are computationally efficient but still suffer from missed detections and unstable localization when dealing with elongated cracks, weak-texture damage and complex pavement backgrounds. To address these issues, this study proposes LDA-YOLO11, a lightweight detail-attention road defect detector based on YOLO11n. The model improves the neck feature fusion stage while keeping the original YOLO11n backbone and P3′ , P4 and P5 detection heads unchanged. A P2 -guided Detail-Neck is designed to inject shallow high-resolution detail information into the P3′ feature without adding an independent P2 detection head. Efficient Channel Attention modules are further introduced after the final P3′ and P4 fusion features to recalibrate defect-related channel responses and suppress background interference. Experimental results show that LDA-YOLO11 achieves [email protected] and [email protected]:0.95 values of 0.951 and 0.791, improving YOLO11n by 2.3 and 1.7% points, respectively. On the Jetson Orin NX platform, TensorRT FP16 acceleration reduces the inference latency to 27 ms/image, corresponding to 37.03 FPS, while causing only a 0.2%-point decrease in [email protected]. These results indicate that LDA-YOLO11 demonstrates promising edge deployment feasibility and real-time inference capability for road surface defect detection.
Related
- A lightweight underwater biological object detector with enhanced cross-scale feature interaction
- Structure-preserving lightweight skin lesion segmentation with dual-resolution attention and boundary-aware fusion
- A lightweight deep learning model for real-time in-vehicle driver distraction detection with low-latency inference
- Deep learning-based detection of construction waste in complex scenarios with an improved lightweight algorithm
- TCLNet: A Thermal Contrast-Aware Lightweight Network for infrared object detection
BibTeX
@article{zhou2026ldayolo11,
title = {{LDA-YOLO11: a lightweight detail-attention detector for edge-deployable road surface defect detection}},
author = {Shiwei Zhou and Shujian Wang and Yang-Kwon Jeong and Jingyan Xiang},
journal = {Scientific Reports},
year = {2026},
month = sep,
doi = {10.1038/s41598-026-69844-1},
url = {https://doi.org/10.1038/s41598-026-69844-1},
note = {CC BY},
}
RIS
TY - JOUR
TI - LDA-YOLO11: a lightweight detail-attention detector for edge-deployable road surface defect detection
AU - Shiwei Zhou
AU - Shujian Wang
AU - Yang-Kwon Jeong
AU - Jingyan Xiang
PY - 2026
DA - 2026/09/01
JO - Scientific Reports
DO - 10.1038/s41598-026-69844-1
UR - https://doi.org/10.1038/s41598-026-69844-1
AB - Road surface defects, such as cracks, pavement damage and potholes, shorten road service life and affect driving safety and preventive maintenance. Existing lightweight object detectors are computationally efficient but still suffer from missed detections and unstable localization when dealing with elongated cracks, weak-texture damage and complex pavement backgrounds. To address these issues, this study proposes LDA-YOLO11, a lightweight detail-attention road defect detector based on YOLO11n. The model improves the neck feature fusion stage while keeping the original YOLO11n backbone and P_3^′ , P_4 and P_5 detection heads unchanged. A P_2 -guided Detail-Neck is designed to inject shallow high-resolution detail information into the P_3^′ feature without adding an independent P_2 detection head. Efficient Channel Attention modules are further introduced after the final P_3^′ and P_4 fusion features to recalibrate defect-related channel responses and suppress background interference. Experimental results show that LDA-YOLO11 achieves [email protected] and [email protected]:0.95 values of 0.951 and 0.791, improving YOLO11n by 2.3 and 1.7% points, respectively. On the Jetson Orin NX platform, TensorRT FP16 acceleration reduces the inference latency to 27 ms/image, corresponding to 37.03 FPS, while causing only a 0.2%-point decrease in [email protected]. These results indicate that LDA-YOLO11 demonstrates promising edge deployment feasibility and real-time inference capability for road surface defect detection.
SN - 2045-2322
ER -