Research and applications of artificial neural network in pavement engineering: A state-of-the-art review

Yang, Xu; Guan, Jinchao; Ding, Ling; You, Zhanping; Lee, Vincent C. S.; Hasan, Mohd Rosli Mohd; Cheng, Xiaoyun

doi:10.1016/j.jtte.2021.03.005

Cited by 48 publications

(18 citation statements)

References 130 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This model was selected for its high efficiency, guaranteed by the convolutional single-layer algorithm structure. These assumptions have been also established by research studies on infrastructure assets-related problems demonstrating the effectiveness of ANNs models in this area of application [54][55][56].…”

Section: Methodology 31 the Implemented Approachmentioning

confidence: 96%

Automatic damage detection of bridge joints and road pavements by artificial neural networks ANNs

Gagliardi¹,

Bella

Sansonetti³

et al. 2022

Earth Resources and Environmental Remote Sensing/Gis Applications XIII

View full text Add to dashboard Cite

Monitoring the actual conditions of critical infrastructure assets is a priority for administrators to guarantee high-standard in terms of structural stability, and operational safety and to prevent damages and deterioration. Nowadays, most protocols for assessing road pavement conditions, including viaducts and bridges, are based on visual inspections conducted by specialized operators and, rarely, the local application of ground-based technologies such as IoT Wireless Sensors, laser scanners, Falling Weight Deflectometers, and accelerometers. However, the high costs of maintenance operations and on-site surveys, which involves the partial or total temporary closure of the infrastructure for the duration of the tests, still limit the implementation of these procedures at the network-scale level. Accordingly, the definition of innovative, automatic, and low-cost procedures for continuous monitoring operations, especially for road pavements and concrete bridges, is still an open challenge. In this context, recent developments in Artificial Neural Networks (ANNs) have opened new possibilities in the automatic recognition and geo-localization of damages affecting critical transportation assets, such as bridges and viaducts. This research presents an experimental application for the automatic detection of deteriorated bridge-deck expansion joints and the classification of pavement distresses based on Artificial Neural Networks models. More specifically, the methodology is founded on the latest generation of ANNs algorithms, among which “YOLO v5”. For this purpose, an experimental evaluation is conducted by the acquisition of a large dataset of imageries of road damages and bridge-deck joints, collected by open-source datasets, and the processing by ANNs. Furthermore, a field experiment was conducted in order to collect several videos of bridges, selected as case-studies, by means of an instrumented vehicle, through a High-Definition resolution camera. The acquired images were post-processed and implemented for the training phases of the ANNs and the verification phases of the developed model. The application of the presented approach provides a reliable, affordable, and promising methodology for the automatic identification and localization of pavement distresses and bridge-deck joints damages, to be rapidly managed and decisively actioned by administrative authorities and asset owners, giving crucial information that could be implemented for the prioritisation of maintenance activities within Asset Management Systems. The findings of this study demonstrate that ANNs approaches and Deep Learning algorithms, can be applied to complement Non-Destructive Remote Sensing technologies (e.g., satellite radar interferometry, Laser Scanners), localizing automatically the pavement and infrastructure damages, paving the way for integrated approaches in the smart monitoring of infrastructure assets.

show abstract

Section: Methodology 31 the Implemented Approachmentioning

confidence: 96%

Automatic damage detection of bridge joints and road pavements by artificial neural networks ANNs

Gagliardi¹,

Bella

Sansonetti³

et al. 2022

Earth Resources and Environmental Remote Sensing/Gis Applications XIII

View full text Add to dashboard Cite

show abstract

“…In the middle of the 20th century, there were attempts to simulate the human brain that finally arrived at something now known as the Artificial Neural Network (ANN) [7]. ANN is a generalized decision-making system, and it is a base for processing the brain's essential elements functions [8].…”

Section: Introductionmentioning

confidence: 99%

Evolving and training of Neural Network to Play DAMA Board Game Using NEAT Algorithm

Qader

Jihad

Baker³

2022

IJCAI

View full text Add to dashboard Cite

Neuroevolutionary algorithms, such as NeuroEvolution of Augmenting Topologies (NEAT) in Machine Learning (ML) methods, are utilized for training and playing computer games due to increased research in Artificial Intelligence (AI) field. NEAT is a genetic algorithm for the generation of evolving artificial neural networks. In this paper, a new study is presented. A Dama board game is designed, and the NEAT algorithm is implemented to develop and train the populations of neural networks for playing the game efficiently. Different inputs and outputs for the network are used, and various network sizes are tried for the game to reach or pass the human level. This paper aims to make a neural network that plays a Dama game like humans or is close by training different neural networks for many generations. The experimental results show that neural networks have been trained for several thousands of generations, and they have played more than one million games. It is concluded that using more input to handle information is better for the learning process. It is also found that a set of values for NEAT parameters is suitable for extensive neural networks like those used in this paper. Povzetek: je predstavljena nova študija. Zasnovana je tradicionalna igra, imenovana "Dama", in implementiran je algoritem NEAT za razvoj in usposabljanje populacij nevronskih omrežij, da igrajo igro, prvič, na način, ki presega človeške zmožnosti.

show abstract

“…People's personal preferences for genres of musical compositions also drive service providers to provide more accurate music classification results. The traditional way of understanding music and performing recursive classification mainly by professionals is unable to cope with the massive amount of music, and it is imperative to use computer programs to automatically classify music genres [ 2 – 5 ]. In the task of music genre classification, many classical machine learning methods have achieved good results on standard datasets, but these methods make extensive use of recursive features designed by domain experts, and the threshold for non-domain experts is high.…”

Section: Introductionmentioning

confidence: 99%

Design of Semiautomatic Digital Creation System for Electronic Music Based on Recurrent Neural Network

Duan

Wang²

2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

Semiautomated digital creation is increasingly important in the manipulation of electronic music. How to realize the learning of local effective features of audio data is a difficult point in the current research field. Based on recurrent neural network theory, this paper designs a semiautomatic digital creation system for electronic music for digital manipulation and genre classification. The recurrent neural network improves the transmission of electronic music information between the input and output of the network by adopting dense connections consistent with DenseNet and adopts an inception-like structure for the autonomous selection of effective recursive nuclear electronic music categories. In the simulation process, the prediction method based on semiautomatic digital audio clips is also adopted, which pays more attention to the learning of local effective features of audio data, which gives the model the ability to create audio samples of different lengths and improves the model’s support for creative tasks in different scenarios. It includes the determination of the number of neurons, the selection of the function of neurons, the determination of the connection method, and the specific learning algorithm rules, and then the training samples are formed. The experimental results show that the recurrent neural network exhibits powerful feature extraction ability and classification ability of music information. The 10-fold cross-validation on GTZAN dataset and ISMIR2004 dataset has obtained 88.7% and 87.68%, surpassing similar ones. The model has reached a leading level. After further use of the MSD (Million Song Dataset) dataset for pre-semiautomatic training, the model effect has been further greatly improved. The accuracy rate on the dataset has been increased to 91.0% and 89.91%, respectively, which has improved the semiautomatic number and creative advancement.

show abstract

Research and applications of artificial neural network in pavement engineering: A state-of-the-art review

Cited by 48 publications

References 130 publications

Automatic damage detection of bridge joints and road pavements by artificial neural networks ANNs

Automatic damage detection of bridge joints and road pavements by artificial neural networks ANNs

Evolving and training of Neural Network to Play DAMA Board Game Using NEAT Algorithm

Design of Semiautomatic Digital Creation System for Electronic Music Based on Recurrent Neural Network

Contact Info

Product

Resources

About