Crack Detection in Infrastructure Using Transfer Learning, Spatial Attention, and Genetic Algorithm Optimization

Feng Ding

REVIEW 4 major objections 4 minor 27 references

Crack Detection in Infrastructure Using Transfer Learning, Spatial Attention, and Genetic Algorithm Optimization

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that a frozen ResNet50, a spatial attention layer, and a genetic-algorithm-tuned classifier detect cracks with F1 score 0.9983.

desk verdict Incremental CBAM-style attention plus GA-tuned head; the near-perfect numbers are unverifiable because no test split is ever defined, and the GA selects on validation accuracy. read the letter →

arxiv 2411.17140 v1 pith:MS7UWKEX submitted 2024-11-26 cs.CV

Feng Ding This is my paper · ORCID

classification cs.CV

keywords crackdetectiontransferlearningResNet50spatialattentiongeneticalgorithmstructuralhealthmonitoringimageclassificationhyperparameteroptimization

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that crack detection in infrastructure images can be made almost flawless even when labeled data is scarce, by chaining three ideas: frozen ImageNet features from ResNet50, a spatial attention layer that highlights crack-like regions, and a small custom classifier whose layer sizes are selected by a genetic algorithm. On a balanced 6,000-image subset drawn from the Crack Forest and Concrete Crack Images for Classification datasets, the resulting Attention-ResNet50-GA model reports precision 0.9967, recall 1.0, and F1 score 0.9983, surpassing a CNN trained from scratch, plain transfer-learned ResNet50, and ResNet50 with attention only. The practical stake is that a high-accuracy, low-data pipeline could reduce reliance on expensive manual inspection and large annotated datasets.

What carries the argument

The load-bearing mechanism is the three-stage pipeline. First, ResNet50 pretrained on ImageNet is used with frozen weights as a fixed feature extractor. Second, a spatial attention module computes average-pooled and max-pooled channel maps, concatenates them, applies a convolution and sigmoid to produce a per-pixel weight map, and multiplies it back into the feature tensor. Third, a customized fully connected network—whose number of layers (chosen as 4) and neurons per layer (66, 805, 218, 382) are encoded as a chromosome and evolved by a genetic algorithm using roulette-wheel selection, random pairing, crossover, and mutation, with validation accuracy as fitness—turns the attended features into a crack/no-crack decision.

What would settle it

Run the proposed Attention-ResNet50-GA pipeline on a fixed, pre-registered split that holds out, say, 20% of the 6,000 images before any GA run, training the GA only on the training fold with its own validation fold, and report metrics on the untouched test fold; if the F1 score drops substantially from 0.9983, the near-perfect result is an artifact of selection on the validation set. Alternatively, apply the trained model to the full CCICD test set or to the 329 Crack Forest images and check whether precision and recall remain at the reported level.

Watch

Extended reading notes

Core claim

The paper's central claim is that the Attention-ResNet50-GA model—a ResNet50 feature extractor kept frozen, a spatial attention map built from channel-wise average and max pooling followed by a sigmoid-activated convolution, and a four-layer fully connected head with 66, 805, 218, and 382 neurons found by the genetic algorithm—detects cracks with near-perfect accuracy. The reported numbers are precision 0.9967, recall 1.0, and F1 score 0.9983 on the described dataset, outperforming all three baselines in Table 1. The paper attributes the gain to the GA-selected head, since the attention-only variant (Att-ResNet50) actually had lower precision than plain transfer-learned ResNet50, showing that attention alone does not help without the customized, optimized classifier.

Load-bearing premise

The reported near-perfect metrics assume the final numbers come from a held-out test set that is disjoint from the validation set used to guide the genetic algorithm and from the images that influenced the balanced-subset selection, but no such split is described.

Editorial extensions

If this is right

If the reported metrics reproduce on a properly held-out test set, near-perfect crack detection is achievable with only a few thousand labeled images because the pretrained feature extractor supplies most of the representation.
The comparison implies that adding a spatial attention layer to a frozen backbone can hurt precision when trained on a small set, but that a GA-tuned classification head can more than compensate.
A deployable version could reuse the same frozen ResNet50 and attention layer for new inspection surfaces by re-running the GA search on a small domain-specific sample.
The GA-selected architecture of 4 layers with 66, 805, 218, and 382 neurons gives practitioners a concrete starting point without repeating the search.
The reported model size of 91.69 MB and 24.03 million parameters is modest enough for edge deployment if the frozen backbone is kept intact.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

One testable extension is to compare the GA-selected head against a random search over the same layer-count and neuron ranges; if random search matches the performance, the gain comes from added capacity rather than from the evolutionary search itself.
Because the GA fitness function is validation accuracy and no test split is described, the near-perfect numbers could reflect selection on the validation set; re-evaluating on a fixed, untouched test split, or on the full CCICD test set, would show whether the claim holds out of sample.
The same pipeline could be evaluated per surface material (asphalt, concrete, masonry) to see whether the attention layer localizes cracks differently across conditions, since the paper's mixed dataset does not separate them.
Once trained, the spatial attention map could be exported as a coarse crack-localization heatmap, giving free localization information alongside the classification label.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Quick take: this is an incremental applied paper that stacks three off-the-shelf ideas — frozen ImageNet ResNet50 features, a CBAM-style spatial attention block, and a GA-searched classifier head — for infrastructure crack detection. The actual novelty is the combination itself, and the empirical centerpiece — precision 0.9967, F1 0.9983 — cannot be evaluated as reported. The paper never states a train/validation/test split. The GA fitness function is explicitly validation accuracy, and no independent test set is described anywhere. If Table 1 comes from the same validation set used for architecture selection, the near-perfect scores are partly self-selected. That is a load-bearing omission, not a style quibble.

Credit where it is due: the spatial attention equations are standard and correctly laid out; the GA encoding, crossover, and mutation are described in enough detail to reimplement; and the limitation section honestly acknowledges that attention alone did not help and that GA is computationally expensive. The choice of ResNet50 because the cited comparison study did not cover it is reasonable and clearly motivated. The literature review is adequate for an applied paper.

Soft spots, in proportion. The missing split is the big one. There are also no seeds, repeated runs, confidence intervals, code, or data release, so the near-perfect result cannot be checked. GA hyperparameters (population size, mutation probabilities, neuron range, generations) are never given. The 6,000-image subset was 'manually selected' from various sources, but the selection criteria and whether selection happened before any split are unclear. On the modeling side, the spatial attention block is essentially CBAM without the channel branch, yet there is no comparison against a standard CBAM-ResNet50 baseline, which would be the natural control. It is also odd that adding attention alone lowers precision (0.914 vs 0.9667) while GA plus attention jumps to 0.9967; the paper's sample-size explanation is hand-wavy.

Citation pattern: fine. The references to transfer learning, attention, and GA in crack detection are appropriate, and the paper does not oversell the literature.

Bottom line: desk reject in current form. The author should release code and data, define the split, report repeated runs with error bars, and add the CBAM baseline. If those are supplied, the result might be worth a second look. As it stands, I would not spend referee time on it.

Referee Report

4 major / 4 minor

Summary. This paper proposes a pipeline for binary crack detection in infrastructure images: a frozen ImageNet-pretrained ResNet50 is used as a feature extractor, the features pass through a spatial attention module, and the resulting representation is classified by a small fully connected network whose layer count and layer widths are optimized by a genetic algorithm. The authors evaluate four configurations (a CNN trained from scratch, plain ResNet50 transfer learning, attention-augmented ResNet50, and attention-augmented ResNet50 with a GA-optimized head) on a balanced subset of 6,000 images drawn from the CFD and CCICD datasets. Table 1 reports precision 0.9967, recall 1.0, and F1 0.9983 for the proposed Attention-ResNet50-GA model, which the authors interpret as near-perfect and superior crack detection. The paper does not describe an independent test set, repeated runs, error bars, or code/data availability.

Significance. The proposed combination of frozen pretrained features, spatial attention, and evolutionary search over the classifier-head architecture is a reasonable and clearly described idea, and the paper documents the GA representation, crossover, and mutation operators in more detail than is common. If the reported numbers were measured on a held-out test set, the result would be a useful incremental demonstration, though not a conceptual surprise, since the components are all established. The central empirical claim is not currently supported, however: the manuscript never defines a test set, the GA fitness is explicitly validation accuracy, and no repeated runs or error bars are reported. The absence of code and data also prevents independent checking. Consequently, the paper's main quantitative contribution—the near-perfect precision and F1—cannot be verified or interpreted as evidence of generalization.

major comments (4)

[Case Study / Result (Table 1), with 'Fitness Function' in 'Optimization Using Genetic Algorithm'] The central claim is load-bearing on an independent test set, but the manuscript never describes one. In the Fitness Function subsection, f(individual) is defined as validation accuracy, and Figure 4 shows training and validation curves only. If the numbers in Table 1 were computed on the same validation set that guided the GA, then the high precision and F1 are partially artifacts of selection rather than evidence of generalization. The authors must specify the train/validation/test split, state that the reported metrics are from a test set never used in GA fitness evaluation or model selection, and give the sizes of the three partitions.
[Data Acquisitions and Pre-processing] The construction of the evaluation dataset is not described with enough precision to assess bias. The paper states that a balanced subset of 6,000 images (3,000 per class) was selected from CFD and CCICD 'to mitigate' the homogeneity of CCICD, but it does not state the selection criterion, whether the selection was random or manual, or how the subset was divided among the two sources. If the subset was chosen with knowledge of the labels or after looking at the images, the reported performance can reflect dataset selection rather than model quality. The authors should describe a reproducible sampling procedure and perform the subset selection before any test split, without access to test labels.
[Result (Table 1) and 'Genetic Algorithm Process'] All quantitative claims rest on single-run point estimates. The GA is stochastic—initialization, roulette-wheel selection, random pairing, and mutation all involve randomness—and the metric differences are small on recall (0.9797 vs 1.0). Without multiple seeds, confidence intervals, or a statistical comparison, the conclusion that Attention-ResNet50-GA 'outperforms' the baselines is not established. Please report the mean and standard deviation over repeated runs and, if possible, a paired significance test across the same folds.
[Optimization Using Genetic Algorithm] The GA is a core component, yet the manuscript does not report the values of its key hyperparameters: population size N, mutation probabilities p_add, p_remove, and p_neuron, the allowed neuron range [n_min, n_max], the number of generations, and the convergence threshold. Only the maximum number of layers (5) is stated. These values are necessary both for reproducibility and for judging whether the search was sufficient to support the optimality claim.

minor comments (4)

[Abstract] The abstract contains a typo: 'inaccessability' should be 'inaccessibility'.
[Genetic Algorithm Process / Crossover] The notation P1 = {L(1)_1, ..., L(1)_n1} is confusing because L is used both for the number of layers and for the layer sizes; please use distinct symbols, e.g., l_i for the number of neurons in layer i.
[Result] The phrase 'with 66 , 805, 218, 382 being the number of the neurons in each layer' is ambiguous; write the architecture as [66, 805, 218, 382] or use a clearer delimiter so the per-layer sizes are unambiguous.
[Data Acquisitions and Pre-processing] Please add a data availability statement or specify where the code and the exact subset of images can be obtained; currently the experiments cannot be reproduced.

Circularity Check

1 steps flagged · score 6.0 of 10

The reported near-perfect precision/F1 is the GA's validation-set fitness function, not an independently tested prediction.

fitted input called prediction [Methodology / Genetic Algorithm Process (Fitness Function); Case Study / Result (Table 1, Figure 4)]
"In the GA, the fitness of each individual was determined based on the model's classification accuracy on the validation set. ... f(individual) = validation accuracy. ... despite having more parameters, the Attention-ResNet50-GA model achieves near-perfect results, with a precision score of 0.9967."

The GA searches over customized-network architectures (number of layers and neurons) by maximizing validation accuracy. The paper then presents the selected model's precision and F1 as evidence of superior crack detection, but it never defines or reports an independent test split; Figure 4 shows only training and validation curves, and the Data section describes only selection of a 6,000-image balanced subset. Under the paper's own description, the reported near-perfect metrics are the same objective that drove architecture selection, so the 'prediction' of generalization is statistically forced by the optimizer rather than demonstrated on unseen data.

full rationale

The only circular step is the evaluation protocol: the GA fitness is explicitly validation accuracy, and no test set is defined anywhere in the paper. If Table 1's metrics are computed on the same validation set used for architecture selection, the central claim of 0.9967 precision / 0.9983 F1 is a restatement of the optimized objective, not an independent result. The remaining methodology (frozen ResNet50 features, CBAM-style spatial attention, GA crossover/mutation) is standard and internally consistent; there are no self-citations or imported uniqueness theorems. However, because the paper's headline result cannot be distinguished from its selection criterion, the circularity score is 6 rather than 0-2.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The central claim rests on several unstated choices: the GA parameters, the architecture search fitness signal, the manual data subset, and the implicit held-out test set. No new physical or conceptual entities are introduced; all components are existing techniques. The absence of a described test split is the most consequential unstated assumption.

free parameters (7)

GA population size N
The initialization step defines N individuals but N is never specified; the search result depends on it.
GA mutation probabilities padd, premove, pneuron
These probabilities are defined symbolically but no values are given; the final architecture depends on them.
Neuron range nmin, nmax
Mutation picks a random value from [Nmin, Nmax], but the bounds are not reported.
GA generations and convergence threshold
The termination condition is 'a specified number of generations or until convergence,' with no numbers provided.
GA-selected classifier architecture = 4 layers with 66, 805, 218, 382 neurons
This architecture is fitted to maximize validation accuracy; the paper's near-perfect metrics are reported for this chosen architecture.
Balanced subset of 6,000 images = 6,000 images, 3,000 per class
The subset size and composition are chosen by the author after noting CCICD homogeneity; this affects every reported metric and the subset is not released.
Training hyperparameters
Epochs, batch size, optimizer, learning rate, and loss function are not reported, so the result cannot be reproduced.

assumptions (6)

standard math Convolution, pooling, sigmoid, and backpropagation behave as standard in deep learning.
The spatial attention and classifier rely on standard deep learning operations.
domain assumption ImageNet-pretrained ResNet50 features transfer to crack detection.
Stated in the Methodology section; not demonstrated by an ablation comparing frozen versus fine-tuned backbones or other pretrained models.
domain assumption Labels in CFD and CCICD are correct and the selected 6,000-image subset is representative.
No annotation quality check or diversity measure is provided.
domain assumption Resizing all images to 224x224 preserves crack-relevant details.
Images are downsampled from 480x320 and 227x227 resolutions; no experiments vary resolution.
ad hoc to paper Validation accuracy is a sufficient fitness signal for architecture search, and the GA-selected architecture generalizes.
The GA uses validation accuracy as fitness, and the paper reports final performance without stating whether the test set is disjoint from the validation set.
ad hoc to paper Selecting 6,000 images from the homogeneous CCICD and CFD yields a diverse, balanced benchmark.
The selection is described qualitatively; no algorithm or diversity metric is given.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Crack Detection in Infrastructure Using Transfer Learning, Spatial Attention, and Genetic Algorithm Optimization." pith.science (2026). https://pith.science/paper/MS7UWKEX

@misc{pith2026241117140,
  author       = {Pith},
  title        = {Pith review of: Crack Detection in Infrastructure Using Transfer Learning, Spatial Attention, and Genetic Algorithm Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MS7UWKEX}},
  note         = {Machine review of arXiv:2411.17140}
}

read the original abstract

Crack detection plays a pivotal role in the maintenance and safety of infrastructure, including roads, bridges, and buildings, as timely identification of structural damage can prevent accidents and reduce costly repairs. Traditionally, manual inspection has been the norm, but it is labor-intensive, subjective, and hazardous. This paper introduces an advanced approach for crack detection in infrastructure using deep learning, leveraging transfer learning, spatial attention mechanisms, and genetic algorithm(GA) optimization. To address the challenge of the inaccessability of large amount of data, we employ ResNet50 as a pre-trained model, utilizing its strong feature extraction capabilities while reducing the need for extensive training datasets. We enhance the model with a spatial attention layer as well as a customized neural network which architecture was fine-tuned using GA. A comprehensive case study demonstrates the effectiveness of the proposed Attention-ResNet50-GA model, achieving a precision of 0.9967 and an F1 score of 0.9983, outperforming conventional methods. The results highlight the model's ability to accurately detect cracks in various conditions, making it highly suitable for real-world applications where large annotated datasets are scarce.

Figures

Figures reproduced from arXiv: 2411.17140 by the authors.

**Figure 2.** Demonstration of training set images Study Design In this study, we utilize ResNet50 as our transfer learning base model. Previous work by Katsigiannis et al. [4] comprehensively evaluated various pre-trained models, including VGG16, VGG19, MobileNetV2, InceptionResNetV2, InceptionV3, and Xception. We specifically chose ResNet50 to explore its unique performance characteristics since it was not covered in their ana… view at source ↗

**Figure 3.** Barplot of model performance combines the attention mechanism with a customized neural network optimized using a genetic algorithm (GA). The GA algorithm identified 4 layers for the customized neural network, with 66, 805, 218, 382 being the number of the neurons in each layer, and despite having more parameters, the Attention-ResNet50-GA model achieves near-perfect results, with a precision score of 0.9967 [PITH_… view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: The variation of loss and accuracy in the training and validation [PITH_FULL_IMAGE:figures/full_fig_p014_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 26 canonical work pages

[1]

K., Lin, J

Ham, Y., Han, K. K., Lin, J. J., and Golparvar-Fard, M. (2016) Visual monitoring of civil infrastructure systems via camera-equipped Unmanned Aerial Vehicles (UA Vs): a review of related works. Visualization in Engi- neering, 4(1), 1–8

work page 2016
[2]

A., Zhan, Y., and Liu, J

Ye, W., Ren, J., Lu, C., Zhang, A. A., Zhan, Y., and Liu, J. (2024) In- telligent detection of fastener defects in ballastless tracks based on deep learning. Automation in Construction, 159, 105280

work page 2024
[3]

B., and Ramana, G

Mishra, M., Louren¸ co, P. B., and Ramana, G. (2022) Structural health monitoring of civil engineering structures by using the internet of things: A review. Journal of Building Engineering, 48, 103954

work page 2022
[4]

(2023) Deep learning for crack detection on masonry fa¸ cades using limited data and transfer learning

Katsigiannis, S., Seyedzadeh, S., Agapiou, A., and Ramzan, N. (2023) Deep learning for crack detection on masonry fa¸ cades using limited data and transfer learning. Journal of Building Engineering, 76, 107105

work page 2023
[5]

J., and Faria, D

Hussain, M., Bird, J. J., and Faria, D. R. (2019) A Study on CNN Transfer Learning for Image Classification. In Lotfi, A., Bouchachia, H., Gegov, A., Langensiepen, C., and McGinnity, M., (eds.), Advances in Computational Intelligence Systems , Cham: Springer International Publishing pp. 191– 202

work page 2019
[6]

(2021) Attention U-Net with Feature Fusion Module for Robust Defect Detection

Xiong, Y.-J., Gao, Y.-B., Wu, H., and Yao, Y. (2021) Attention U-Net with Feature Fusion Module for Robust Defect Detection. Journal of Circuits, Systems and Computers, 30(15), 2150272

work page 2021
[7]

H., and Hutter, F

Elsken, T., Metzen, J. H., and Hutter, F. Neural Architecture Search: A Survey. (2019)

work page 2019
[8]

(2021) Automated crack segmentation in close-range building fa¸ cade inspection images using deep learning techniques

Chen, K., Reichard, G., Xu, X., and Akanmu, A. (2021) Automated crack segmentation in close-range building fa¸ cade inspection images using deep learning techniques. Journal of Building Engineering, 43, 102913

work page 2021

Show all 27 references

[9]

and Tsai, Y

Hsieh, Y.-A. and Tsai, Y. (09, 2020) Machine Learning for Crack Detection: Review and Model Performance Comparison.Journal of Computing in Civil Engineering, 34, 04020038. 15

2020
[10]

(2018) Autonomous Structural Visual Inspection Using Region-Based Deep Learning for Detecting Multiple Damage Types

Cha, Y.-J., Choi, W., Suh, G., Mahmoudkhani, S., and B¨ uy¨ uk¨ ozt¨ urk, O. (2018) Autonomous Structural Visual Inspection Using Region-Based Deep Learning for Detecting Multiple Damage Types. Computer-Aided Civil and Infrastructure Engineering, 33(9), 731–747

2018
[11]

Zhang, A., Wang, K. C. P., Fei, Y., Liu, Y., Chen, C., Yang, G., Li, J. Q., Yang, E., and Qiu, S. (2019) Automated Pixel-Level Pavement Crack Detection on 3D Asphalt Surfaces with a Recurrent Neural Network. Computer-Aided Civil and Infrastructure Engineering, 34(3), 213–229

2019
[12]

(2017) Deep Learning-Based Crack Damage Detection Using Convolutional Neural Networks.Computer- Aided Civil and Infrastructure Engineering, 32(5), 361–378

Cha, Y.-J., Choi, W., and B¨ uy¨ uk¨ ozt¨ urk, O. (2017) Deep Learning-Based Crack Damage Detection Using Convolutional Neural Networks.Computer- Aided Civil and Infrastructure Engineering, 32(5), 361–378

2017
[13]

and G¨ ul, M

Mei, Q. and G¨ ul, M. (2020) A cost effective solution for pavement crack in- spection using cameras and deep neural networks. Construction and Build- ing Materials, 256, 119397

2020
[14]

J., and Maguire, M

Dorafshan, S., Thomas, R. J., and Maguire, M. (2018) Comparison of Deep Convolutional Neural Networks and Edge Detectors for Image-Based Crack Detection in Concrete. Construction and Building Materials, 186, 1031– 1045

2018
[15]

(2019) Automatic Bridge Crack Detection Using a Convolutional Neural Network

Xu, H., Su, X., Wang, Y., Cai, H., Cui, K., and Chen, X. (2019) Automatic Bridge Crack Detection Using a Convolutional Neural Network. Applied Sciences, 9(14)

2019
[16]

and Song, J

Zhu, J. and Song, J. (2020) An Intelligent Classification Model for Surface Defects on Cement Concrete Bridges. Applied Sciences, 10(3)

2020
[17]

K., Choudhary, A., and Agrawal, A

Gopalakrishnan, K., Khaitan, S. K., Choudhary, A., and Agrawal, A. (2017) Deep Convolutional Neural Networks with transfer learning for com- puter vision-based data-driven pavement distress detection. Construction and Building Materials, 157, 322–330

2017
[18]

D., and Zhang, B

Zhang, K., Cheng, H. D., and Zhang, B. (2018) Unified Approach to Pavement Crack and Sealed Crack Detection Using Preclassification Based on Transfer Learning. Journal of Computing in Civil Engineering, 32(2), 04018001

2018
[19]

Kharthik, K. S. B., Onyema, E. M., Mallik, S., Prasad, B. V. V. S., Qin, H., Selvi, C., and Sikha, O. K. (2024) Transfer learned deep feature based crack detection using support vector machine: a comparative study. Scientific Reports, 14

2024
[20]

and Liu, Y.-C

Luo, J.-Y. and Liu, Y.-C. (2024) Adaptive and Explainable Deep Learning- Based Rapid Identification of Architectural Cracks. IEEE Access, 12, 111741–111751. 16

2024
[21]

Crack detection based on attention mechanism with YOLOv5

Lan, M.-L., Yang, D., Zhou, S.-X., and Ding, Y. Crack detection based on attention mechanism with YOLOv5. Engineering Reports, n/a(n/a), e12899
[22]

(2023) Online monitoring of crack dy- namic development using attention-based deep networks

Chen, W., He, Z., and Zhang, J. (2023) Online monitoring of crack dy- namic development using attention-based deep networks. Automation in Construction, 154, 105022

2023
[23]

and Zhu, H

Yu, T. and Zhu, H. (2020) Hyper-Parameter Optimization: A Review of Algorithms and Applications. CoRR, abs/2003.05689

2020 arXiv
[24]

D., Bello-Salau, H., Umoh, I

Raji, I. D., Bello-Salau, H., Umoh, I. J., Onumanyi, A. J., Adegboye, M. A., and Salawudeen, A. T. (2022) Simple Deterministic Selection-Based Ge- netic Algorithm for Hyperparameter Tuning of Machine Learning Models. Applied Sciences, 12(3)

2022
[25]

(2022) Deep Learning of Neural Networks Using Genetic Algorithms

Lienkov, S., Shvorov, S., Sieliukov, O., Tolok, I., Lytvynenko, N., and Davydenko, T. (2022) Deep Learning of Neural Networks Using Genetic Algorithms. In MoMLeT+DS

2022
[26]

(2016) Automatic road crack detection using random structured forests

Shi, Y., Cui, L., Qi, Z., Meng, F., and Chen, Z. (2016) Automatic road crack detection using random structured forests. IEEE Transactions on Intelligent Transportation Systems, 17(12), 3434–3445

2016
[27]

C ¸ a˘ glar Fırat¨Ozgenel Concrete Crack Images for Classification. (2019). 17

2019

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

K., Lin, J

Ham, Y., Han, K. K., Lin, J. J., and Golparvar-Fard, M. (2016) Visual monitoring of civil infrastructure systems via camera-equipped Unmanned Aerial Vehicles (UA Vs): a review of related works. Visualization in Engi- neering, 4(1), 1–8

work page 2016

[2] [2]

A., Zhan, Y., and Liu, J

Ye, W., Ren, J., Lu, C., Zhang, A. A., Zhan, Y., and Liu, J. (2024) In- telligent detection of fastener defects in ballastless tracks based on deep learning. Automation in Construction, 159, 105280

work page 2024

[3] [3]

B., and Ramana, G

Mishra, M., Louren¸ co, P. B., and Ramana, G. (2022) Structural health monitoring of civil engineering structures by using the internet of things: A review. Journal of Building Engineering, 48, 103954

work page 2022

[4] [4]

(2023) Deep learning for crack detection on masonry fa¸ cades using limited data and transfer learning

Katsigiannis, S., Seyedzadeh, S., Agapiou, A., and Ramzan, N. (2023) Deep learning for crack detection on masonry fa¸ cades using limited data and transfer learning. Journal of Building Engineering, 76, 107105

work page 2023

[5] [5]

J., and Faria, D

Hussain, M., Bird, J. J., and Faria, D. R. (2019) A Study on CNN Transfer Learning for Image Classification. In Lotfi, A., Bouchachia, H., Gegov, A., Langensiepen, C., and McGinnity, M., (eds.), Advances in Computational Intelligence Systems , Cham: Springer International Publishing pp. 191– 202

work page 2019

[6] [6]

(2021) Attention U-Net with Feature Fusion Module for Robust Defect Detection

Xiong, Y.-J., Gao, Y.-B., Wu, H., and Yao, Y. (2021) Attention U-Net with Feature Fusion Module for Robust Defect Detection. Journal of Circuits, Systems and Computers, 30(15), 2150272

work page 2021

[7] [7]

H., and Hutter, F

Elsken, T., Metzen, J. H., and Hutter, F. Neural Architecture Search: A Survey. (2019)

work page 2019

[8] [8]

(2021) Automated crack segmentation in close-range building fa¸ cade inspection images using deep learning techniques

Chen, K., Reichard, G., Xu, X., and Akanmu, A. (2021) Automated crack segmentation in close-range building fa¸ cade inspection images using deep learning techniques. Journal of Building Engineering, 43, 102913

work page 2021