1
|
Zeng J, Zhong H. YOLOv8-PD: an improved road damage detection algorithm based on YOLOv8n model. Sci Rep 2024; 14:12052. [PMID: 38802524 PMCID: PMC11130172 DOI: 10.1038/s41598-024-62933-z] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/01/2024] [Accepted: 05/22/2024] [Indexed: 05/29/2024] Open
Abstract
Road damage detection is an crucial task to ensure road safety. To tackle the issues of poor performance on multi-scale pavement distresses and high costs in detection task, this paper presents an improved lightweight road damage detection algorithm based on YOLOv8n, named YOLOv8-PD (pavement distress). Firstly, a BOT module that can extract global information of road damage images is proposed to adapt to the large-span features of crack objects. Secondly, the introduction of the large separable kernel attention (LKSA) mechanism enhances the detection accuracy of the algorithm. Then, a C2fGhost block is constructed in the neck network to strengthen the feature extraction of complex road damages while reducing the computational load. Furthermore, we introduced lightweight shared convolution detection head (LSCD-Head) to improve feature expressiveness and reduce the number of parameters. Finally, extensive experiments on the RDD2022 dataset yield a model with parametric and computational quantities of 2.3M and 6.1 GFLOPs, which are only 74.1% and 74.3% of the baseline, and the mAP reaches an improvement of 1.4 percentage points from the baseline. In addition, experimental results on the RoadDamage dataset show that the mAP increased by 4.2% and this algorithm has good robustness. This method can provide a reference for the automatic detection method of pavement distress.
Collapse
Affiliation(s)
- Jiayi Zeng
- College of information and Network Safety, People's Public Security University of China, Beijing, 100038, China
| | - Han Zhong
- College of information and Network Safety, People's Public Security University of China, Beijing, 100038, China.
| |
Collapse
|
2
|
Deng Y, Li F, Zhou S, Zhang S, Yang Y, Zhang Q, Li Y. Use of recurrent neural networks considering maintenance to predict urban road performance in Beijing, China. PHILOSOPHICAL TRANSACTIONS. SERIES A, MATHEMATICAL, PHYSICAL, AND ENGINEERING SCIENCES 2023; 381:20220175. [PMID: 37454686 DOI: 10.1098/rsta.2022.0175] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 09/19/2022] [Accepted: 02/02/2023] [Indexed: 07/18/2023]
Abstract
A correct understanding of the pavement performance change law forms the premise of the scientific formulation of maintenance decisions. This paper aims to develop a predictive model taking into account the costs of different types of maintenance works that reflects the continuous true usage performance of the pavement. The model proposed in this study was trained on a dataset containing five-year maintenance work data on urban roads in Beijing with pavement performance indicators for the corresponding years. The same roads were matched and combined to obtain a set of sequences of pavement performance changes with the features of the current year; with the recurrent-neural-network-based long short-term memory (LSTM) network and gate recurrent unit (GRU) network, the prediction accuracy of highway pavement performance on the test set was significantly increased. The prediction result indicates that the generalization ability of the improved recurrent neural network model is satisfactory, with the R2 achieving 0.936, and of the two models the GRU model is more efficient, with an accuracy that reaches almost the same level as LSTM but with the training convergence time reduced to 25 s. This study demonstrates that data generated by the work of maintenance units can be used effectively in the prediction of pavement performance. This article is part of the theme issue 'Artificial intelligence in failure analysis of transportation infrastructure and materials'.
Collapse
Affiliation(s)
- Yutong Deng
- Beijing Key Laboratory for Cooperative Vehicle Infrastructure Systems and Safety Control, School of Transportation Science and Engineering, Beihang University, Beijing 100191, People's Republic of China
| | - Feng Li
- Beijing Key Laboratory for Cooperative Vehicle Infrastructure Systems and Safety Control, School of Transportation Science and Engineering, Beihang University, Beijing 100191, People's Republic of China
| | - Siqi Zhou
- Beijing Key Laboratory for Cooperative Vehicle Infrastructure Systems and Safety Control, School of Transportation Science and Engineering, Beihang University, Beijing 100191, People's Republic of China
| | - Song Zhang
- Beijing Municipal Road and Bridge Management and Maintenance Group Co. LTD, Beijing 100097, People's Republic of China
| | - Yang Yang
- Beijing Urban Road Maintenance and Management Center,Beijing 100053, People's Republic of China
| | - Qiang Zhang
- Beijing Municipal Road and Bridge Management and Maintenance Group Co. LTD, Beijing 100097, People's Republic of China
| | - Yanfei Li
- Beijing Municipal Road and Bridge Management and Maintenance Group Co. LTD, Beijing 100097, People's Republic of China
| |
Collapse
|
3
|
Prediction of International Roughness Index Based on Stacking Fusion Model. SUSTAINABILITY 2022. [DOI: 10.3390/su14126949] [Citation(s) in RCA: 8] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 02/04/2023]
Abstract
Pavement performance prediction is necessary for road maintenance and repair (M&R) management and plans. The accuracy of performance prediction affects the allocation of maintenance funds. The international roughness index (IRI) is essential for evaluating pavement performance. In this study, using the road pavement data of LTPP (Long-Term Pavement Performance), we screened the feature parameters used for IRI prediction using the mean decrease impurity (MDI) based on random forest (RF). The effectiveness of this feature selection method was proven suitable. The prediction accuracies of four promising prediction models were compared, including Gradient Boosting Decision Tree (GBDT), eXtreme Gradient Boosting (XGBoost), support vector machine (SVM), and multiple linear regression (MLR). The two integrated learning algorithms, GBDT and XGBoost, performed well in prediction. GBDT performs best with the lowest root mean square error (RMSE) of 0.096 and the lowest mean absolute error (MAE) of 6.2% and the coefficient of determination (R2) reaching 0.974. However, the prediction accuracy varies in numerical intervals, with some deviations. The stacking fusion model with a powerful generalization capability is proposed to build a new prediction model using GBDT and XGBoost as the base learners and bagging as the meta-learners. The R2, RMSE, and MAE of the stacking fusion model are 0.996, 0.040, and 1.3%, which further improves the prediction accuracy and verifies the superiority of this fusion model in pavement performance prediction. Besides, the prediction accuracy is generally consistent across different numerical intervals.
Collapse
|
4
|
Arabi S, Sharma A, Reyes M, Hamann C, Peek-Asa C. Farm Vehicle Following Distance Estimation Using Deep Learning and Monocular Camera Images. SENSORS 2022; 22:s22072736. [PMID: 35408350 PMCID: PMC9003299 DOI: 10.3390/s22072736] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 01/26/2022] [Revised: 03/11/2022] [Accepted: 03/21/2022] [Indexed: 02/04/2023]
Abstract
This paper presents a comprehensive solution for distance estimation of the following vehicle solely based on visual data from a low-resolution monocular camera. To this end, a pair of vehicles were instrumented with real-time kinematic (RTK) GPS, and the lead vehicle was equipped with custom devices that recorded video of the following vehicle. Forty trials were recorded with a sedan as the following vehicle, and then the procedure was repeated with a pickup truck in the following position. Vehicle detection was then conducted by employing a deep-learning-based framework on the video footage. Finally, the outputs of the detection were used for following distance estimation. In this study, three main methods for distance estimation were considered and compared: linear regression model, pinhole model, and artificial neural network (ANN). RTK GPS was used as the ground truth for distance estimation. The output of this study can contribute to the methodological base for further understanding of driver following behavior with a long-term goal of reducing rear-end collisions.
Collapse
Affiliation(s)
- Saeed Arabi
- Department of Civil, Construction, and Environmental Engineering, Iowa State University, Ames, IA 50011, USA;
- Correspondence:
| | - Anuj Sharma
- Department of Civil, Construction, and Environmental Engineering, Iowa State University, Ames, IA 50011, USA;
| | - Michelle Reyes
- National Advanced Driving Simulator, University of Iowa, 127 NADS, Iowa City, IA 52242, USA;
| | - Cara Hamann
- Department of Epidemiology, University of Iowa, 145 N Riverside Dr, S449 CPHB, Iowa City, IA 52242, USA;
- Injury Prevention Research Center, University of Iowa, 145 N Riverside Dr, Iowa City, IA 52242, USA;
| | - Corinne Peek-Asa
- Injury Prevention Research Center, University of Iowa, 145 N Riverside Dr, Iowa City, IA 52242, USA;
- Department of Occupational and Environmental Health, University of Iowa, 145 N Riverside Dr, S143 CPHB, Iowa City, IA 52241, USA
- Office of Research Affairs, University of California San Diego, 9500 Gilman Drive, #0043, La Jolla, CA 92093, USA
| |
Collapse
|
5
|
Abstract
The latest advancements in road asphalt materials and construction technologies have increased the difficulty for engineers to select the appropriate pavement design solution with consideration of proper timing for maintenance planning. On the other hand, Building Information Modeling (BIM) tools allow practitioners to efficiently store and manage large amounts of data, supporting decision making in road asphalt pavement design and management. The present work focused on the elaboration of a BIM-based maintenance analysis tool for the specific evaluation of several condition indicators and the selection of proper maintenance solutions designed to include alternative materials and advanced recycling technologies. A traditional BIM workflow was integrated with additional user-defined property sets to investigate the need for maintenance at the present date and predict the degradation curve of the condition indicators through the least square interpolation of time series of data. The analysis tool also provided the selection of available pavement alternatives from a library of designed solutions based on their compliance with project-specific constraints (maximum budget, minimum useful life, and availability of secondary raw materials and in-place recycling technologies). The proposed BIM tool aims to be a practical and dynamic way to integrate maintenance planning considerations into road pavement design, encouraging the use of digital tools in the road industry and ultimately supporting a pavement maintenance decision-making process oriented towards a circular economy.
Collapse
|
6
|
Esfandiari Y, Balu A, Ebrahimi K, Vaidya U, Elia N, Sarkar S. A fast saddle-point dynamical system approach to robust deep learning. Neural Netw 2021; 139:33-44. [PMID: 33677377 DOI: 10.1016/j.neunet.2021.02.021] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/20/2020] [Revised: 02/18/2021] [Accepted: 02/18/2021] [Indexed: 11/29/2022]
Abstract
Recent focus on robustness to adversarial attacks for deep neural networks produced a large variety of algorithms for training robust models. Most of the effective algorithms involve solving the min-max optimization problem for training robust models (min step) under worst-case attacks (max step). However, they often suffer from high computational cost from running several inner maximization iterations (to find an optimal attack) inside every outer minimization iteration. Therefore, it becomes difficult to readily apply such algorithms for moderate to large size real world data sets. To alleviate this, we explore the effectiveness of iterative descent-ascent algorithms where the maximization and minimization steps are executed in an alternate fashion to simultaneously obtain the worst-case attack and the corresponding robust model. Specifically, we propose a novel discrete-time dynamical system-based algorithm that aims to find the saddle point of a min-max optimization problem in the presence of uncertainties. Under the assumptions that the cost function is convex and uncertainties enter concavely in the robust learning problem, we analytically show that our algorithm converges asymptotically to the robust optimal solution under a general adversarial budget constraints as induced by ℓp norm, for 1≤p≤∞. Based on our proposed analysis, we devise a fast robust training algorithm for deep neural networks. Although such training involves highly non-convex robust optimization problems, empirical results show that the algorithm can achieve significant robustness compared to other state-of-the-art robust models on benchmark data sets.
Collapse
Affiliation(s)
| | - Aditya Balu
- Iowa State University, United States of America
| | | | | | - Nicola Elia
- University of Minnesota, United States of America
| | | |
Collapse
|