Benefits of ensemble models in road pavement cracking classification

Rodríguez-Lozano, Francisco Javier; León‐García, Fernando; Gámez-Granados, Juan Carlos; Palomares, José Manuel; Olivares, Joaquín

doi:10.1111/mice.12543

Cited by 29 publications

(22 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A notable number of researchers have used Artificial Neural Networks (ANN) to predict pavement performance (15)(16)(17)(18)(19). Other ML algorithms used in modeling pavement performance include Decision Trees (20,21), Ensemble Trees (22,23), Random Forest (RF) (24)(25)(26), Support Vector Machines (SVM) (20,(27)(28)(29), and Recurrent Neural Networks (30). The description is focused on ANN, RF, and SVM since these algorithms are the ones most frequently used in modeling IRI.…”

Section: Machine Learning Algorithms For Pavement Performance Modelingmentioning

confidence: 99%

Performance of Machine Learning Algorithms in Predicting the Pavement International Roughness Index

Bashar

Torres-Machí

2021

Transportation Research Record: Journal of the Transportation R

View full text Add to dashboard Cite

Significant research efforts have documented the capabilities of machine learning (ML) algorithms to model pavement performance. Several challenges, however, limit the implementation of ML by practitioners and transportation agencies. One of these challenges is related to the high variability in the performance of ML models as reported by different studies and the lack of quantitative evidence supporting the true effectiveness of these techniques. The objective of this paper is twofold: to assess the overall performance of traditional and ML techniques used to predict pavement condition, and to provide guidance on the optimal architecture and minimum sample size required to develop these models. This paper analyzes three ML algorithms commonly used to predict International Roughness Index (IRI)—Artificial Neural Network (ANN), Random Forest (RF), and Support Vector Machine (SVM)—and compares their performance to traditional techniques. An inverse variance heterogeneity based meta-analysis is performed on 20 studies conducted between 2001 and 2020. The results indicate that ML algorithms capture on average 15.6% more variability than traditional techniques. RF is the most accurate technique with an overall performance value of 0.995. ANN is also identified as a highly effective technique that has been widely used and provides accurate predictions with both small and large sample sizes. For ANN algorithms, a single hidden layer with nodes equal to 0.3–2 times the number of input features is found to be sufficient in predicting pavement deterioration. A minimum sample size equal to 50 times the number of input variables is recommend to model pavement deterioration using ML.

show abstract

Section: Machine Learning Algorithms For Pavement Performance Modelingmentioning

confidence: 99%

Performance of Machine Learning Algorithms in Predicting the Pavement International Roughness Index

Bashar

Torres-Machí

2021

Transportation Research Record: Journal of the Transportation R

View full text Add to dashboard Cite

show abstract

“…Besides ANN, decision trees (Hoang & Nguyen, 2019) and support vector machines (SVM) (Wang et al., 2017) were also used. Rodriguez‐Lozano, León‐García, Gámez‐Granados, Palomares, and Olivares (2020) proposed an ensemble model for crack type classification. In recent years, research efforts started to pay attention to DL models for crack type classification.…”

Section: Literature Reviewmentioning

confidence: 99%

Convolutional neural network for automated classification of jointed plain concrete pavement conditions

Hsieh

Yang

Tsai

2020

Computer aided Civil Eng

View full text Add to dashboard Cite

The detailed monitoring of jointed plain concrete pavement (JPCP) slab condition is essential for cost-effective JPCP maintenance and rehabilitation. However, existing visual inspection practices for detailed slab condition classification are time-consuming and labor-intensive. In this paper, we proposed an automated JPCP slab condition classification model based on convolutional neural networks (ConvNets), which is the first to perform multi-label classification on the JPCP slab condition based on both crack types and severity levels. To handle the different scales between JPCP slab condition states, the model includes a novel global context block with atrous spatial pyramid pooling, denoted as a GC-ASPP block.The block can be flexibly applied to any ConvNets to effectively model the global context of images with the extraction of multiscale image features. The proposed model was evaluated using real-world 3D JPCP surface data. With the GC-ASPP block, our best model achieved an average precision of 85.42% on multi-label slab condition classification.

show abstract

“…Pavement distress significantly endangers traffic safety (Chen et al, 2019;Zeng et al, 2014) and reduces road life. Automated pavement damage detection technology (Liu et al, 2020;Rodriguez-Lozano et al, 2020;Ye et al, 2020; plays a pivotal role in the rapid evaluation of road conditions and maintenances (Huang et al, 2014). Early repair of road defects should be conducted to avoid further degradation (Ahmed et al, 2017;L.…”

Section: Introductionmentioning

confidence: 99%

Cross‐scene pavement distress detection by a novel transfer learning framework

Che

Liu

et al. 2021

Computer aided Civil Eng

View full text Add to dashboard Cite

Deep learning has achieved promising results in pavement distress detection. However, the training model's effectiveness varies according to the data and scenarios acquired by different camera types and their installation positions. It is time consuming and labor intensive to recollect labeled data and retrain a new model every time the scene changes. In this paper, we propose a transfer learning pipeline to address this problem, which enables a distress detection model to be applied to other untrained scenarios. The framework consists of two main components: data transfer and model transfer. The former trains a generative adversarial network to transfer existing image data into a new scene style. Then, attentive CutMix and image melding are applied to insert distress annotations to synthesize the new scene's labeled data. After data expansion, the latter step transfers the feature extracted by the existing model to the detection application of the new scene through domain adaptation. The effects of varying degrees of knowledge transfer are also discussed. The proposed method is evaluated on two data sets from two different scenes with more than 40,000 images totally. This method can reduce the demand for training data by at least 25% when the model is applied in a new scene. With the same number of training images, the proposed method can improve the model accuracy by 26.55%.

show abstract

Benefits of ensemble models in road pavement cracking classification

Cited by 29 publications

References 62 publications

Performance of Machine Learning Algorithms in Predicting the Pavement International Roughness Index

Performance of Machine Learning Algorithms in Predicting the Pavement International Roughness Index

Convolutional neural network for automated classification of jointed plain concrete pavement conditions

Cross‐scene pavement distress detection by a novel transfer learning framework

Contact Info

Product

Resources

About