Lightweight concrete crack recognition model based on improved MobileNetV3
Abstract This study created the C//Sim attention mechanism employing the parallel connection of the CA attention mechanism and the SimAm attention mechanism to detect cracks in lightweight concrete. MobileNetV3 was improved using the above method, and a lightweight concrete crack recognition model,...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-05-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-00468-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849728328723857408 |
|---|---|
| author | Rui Wang Ruiqi Chen Hao Yan Xinxin Guo |
| author_facet | Rui Wang Ruiqi Chen Hao Yan Xinxin Guo |
| author_sort | Rui Wang |
| collection | DOAJ |
| description | Abstract This study created the C//Sim attention mechanism employing the parallel connection of the CA attention mechanism and the SimAm attention mechanism to detect cracks in lightweight concrete. MobileNetV3 was improved using the above method, and a lightweight concrete crack recognition model, MobileNetV3-C//Sim, was established. To validate the model’s practicality, this paper has been tested on self-built and public datasets. The improved model performs higher accuracy, recall, precision, and F1 values than Mobilenetv3 in both datasets, with increases of 0.44–0.69% and 0.46–0.89% for the binary and multi classification tasks, respectively. For the CA attention mechanism, SimAm attention mechanism, and ablation tests with different combinations of each other showed that the parallel connection combination was superior to the single-type, front-to-back concatenation combination. In noise testing with different attention mechanisms, the C//Sim reduction is the smallest. It is verified to have better noise immunity and robustness. Regarding the number of model parameters, the proposed method involves only 2.90 M, which is 30.17% less than that of MobileNetV3. The method can provide a model reference for further concrete crack lightweight identification research. |
| format | Article |
| id | doaj-art-5fac74a7294344879dfee2ac6a504cd9 |
| institution | DOAJ |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-5fac74a7294344879dfee2ac6a504cd92025-08-20T03:09:35ZengNature PortfolioScientific Reports2045-23222025-05-0115111210.1038/s41598-025-00468-7Lightweight concrete crack recognition model based on improved MobileNetV3Rui Wang0Ruiqi Chen1Hao Yan2Xinxin Guo3College of Engineering, Sichuan Normal UniversityCollege of Engineering, Sichuan Normal UniversityCollege of Engineering, Sichuan Normal UniversityState Key Laboratory of Geohazard Prevention and Geoenvironmental Protection, Chengdu University of TechnologyAbstract This study created the C//Sim attention mechanism employing the parallel connection of the CA attention mechanism and the SimAm attention mechanism to detect cracks in lightweight concrete. MobileNetV3 was improved using the above method, and a lightweight concrete crack recognition model, MobileNetV3-C//Sim, was established. To validate the model’s practicality, this paper has been tested on self-built and public datasets. The improved model performs higher accuracy, recall, precision, and F1 values than Mobilenetv3 in both datasets, with increases of 0.44–0.69% and 0.46–0.89% for the binary and multi classification tasks, respectively. For the CA attention mechanism, SimAm attention mechanism, and ablation tests with different combinations of each other showed that the parallel connection combination was superior to the single-type, front-to-back concatenation combination. In noise testing with different attention mechanisms, the C//Sim reduction is the smallest. It is verified to have better noise immunity and robustness. Regarding the number of model parameters, the proposed method involves only 2.90 M, which is 30.17% less than that of MobileNetV3. The method can provide a model reference for further concrete crack lightweight identification research.https://doi.org/10.1038/s41598-025-00468-7Crack recognitionClassificationMobileNetV3C//Sim attention mechanisms |
| spellingShingle | Rui Wang Ruiqi Chen Hao Yan Xinxin Guo Lightweight concrete crack recognition model based on improved MobileNetV3 Scientific Reports Crack recognition Classification MobileNetV3 C//Sim attention mechanisms |
| title | Lightweight concrete crack recognition model based on improved MobileNetV3 |
| title_full | Lightweight concrete crack recognition model based on improved MobileNetV3 |
| title_fullStr | Lightweight concrete crack recognition model based on improved MobileNetV3 |
| title_full_unstemmed | Lightweight concrete crack recognition model based on improved MobileNetV3 |
| title_short | Lightweight concrete crack recognition model based on improved MobileNetV3 |
| title_sort | lightweight concrete crack recognition model based on improved mobilenetv3 |
| topic | Crack recognition Classification MobileNetV3 C//Sim attention mechanisms |
| url | https://doi.org/10.1038/s41598-025-00468-7 |
| work_keys_str_mv | AT ruiwang lightweightconcretecrackrecognitionmodelbasedonimprovedmobilenetv3 AT ruiqichen lightweightconcretecrackrecognitionmodelbasedonimprovedmobilenetv3 AT haoyan lightweightconcretecrackrecognitionmodelbasedonimprovedmobilenetv3 AT xinxinguo lightweightconcretecrackrecognitionmodelbasedonimprovedmobilenetv3 |