SAM Era: Can It Segment Any Industrial Surface Defects?
Source abstract
Abstract
Segment Anything Model (SAM) is a cutting-edge model that has shown impressive performance in general object segmentation. The birth of the segment anything is a groundbreaking step towards creating a universal intelligent model. Due to its superior performance in general object segmentation, it quickly gained attention and interest. This makes SAM particularly attractive in industrial surface defect segmentation, especially for complex industrial scenes with limited training data. However, its segmentation ability for specific industrial scenes remains unknown. Therefore, in this work, we select three representative and complex industrial surface defect detection scenarios, namely strip steel surface defects, tile surface defects, and rail surface defects, to evaluate the segmentation performance of SAM. Our results show that although SAM has great potential in general object segmentation, it cannot achieve satisfactory performance in complex industrial scenes. Our test results are available at: .
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Industrial Vision Systems and Defect Detection · Advanced Neural Network Applications · Visual Attention and Saliency Detection
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (31)
Show source-provided references
1. A. Kirillov et al., “Segment anything,” arXiv preprint arXiv:2304.02643, 2023.
2. J. Ma and B. Wang, “Segment anything in medical images,” arXiv preprint arXiv:2304.12306, 2023.10.1038/s41467-024-44824-z ↗
3. T. Shaharabany, A. Dahan, R. Giryes, and L. Wolf, “AutoSAM: Adapting SAM to medical images by overloading the prompt encoder,” arXiv preprint arXiv:2306.06370, 2023.
4. S. Gong et al., “3DSAM-adapter: Holistic adaptation of SAM from 2D to 3D for promptable medical image segmentation,” arXiv preprint arXiv:2306.13465, 2023.10.1016/j.media.2024.103324 ↗
5. Y. Gao, W. Xia, D. Hu, and X. Gao, “DeSAM: Decoupling segment anything model for generalizable medical image segmentation,” arXiv preprint arXiv:2306.00499, 2023.10.1007/978-3-031-72390-2_48 ↗
6. W. Lei, X. Wei, X. Zhang, K. Li, and S. Zhang, “MedLSAM: Localize and segment anything model for 3D medical images,” arXiv preprint arXiv:2306.14752, 2023.
7. Zero-shot surface defect recognition with class knowledge graph10.1016/j.aei.2022.101813 ↗
8. Zero shot detection10.1109/TCSVT.2019.2899569 ↗
9. Prompt engineering for zero-shot and few-shot defect detection and classification using a visual-language pretrained model10.1111/mice.12954 ↗
10. Synthetic datasets for deep learning in computer-vision assisted tasks in manufacturing10.1016/j.procir.2021.10.038 ↗
11. A vision-based system for real-time defect detection: A rubber compound part case study10.1016/j.procir.2020.04.159 ↗
12. Cascaded partial decoder for fast and accurate salient object detection
13. Stacked cross refinement network for edge-aware salient object detection
14. Multi-scale interactive network for salient object detection
15. Global context-aware progressive aggregation network for salient object detection
16. Wei · Proc. AAAI. Conf. Artif. Intell. · 2020 · 34 · 12321
17. Y. Sun, G. Chen, T. Zhou, Y. Zhang, and N. Liu, “Context-aware crosslevel fusion network for camouflaged object detection,” arXiv preprint arXiv:2105.12555, 2021.10.24963/ijcai.2021/142 ↗
18. Visual saliency transformer
19. Salient object detection via integrity learning
20. EDN: Salient object detection via extremely-downsampled network10.1109/TIP.2022.3164550 ↗
21. Lightweight salient object detection in optical remote-sensing images via semantic matching and edge alignment
22. EDRNet: Encoder-decoder residual network for salient object detection of strip steel surface defects10.1109/TIM.2020.3002277 ↗
23. Dense attention-guided cascaded network for salient object detection of strip steel surface defects
24. Cross-scale edge purification network for salient object detection of steel defect images10.1016/j.measurement.2022.111429 ↗
25. Saliency detection for strip steel surface defects using multiple constraints and improved texture features10.1016/j.optlaseng.2019.106000 ↗
26. Surface defect saliency of magnetic tile10.1007/s00371-018-1588-5 ↗
27. MCnet: Multiple context information segmentation network of no-service rail surface defects10.1109/TIM.2021.3127641 ↗
28. Structure-measure: A new way to evaluate foreground maps
29. D. P. Fan, C. Gong, Y. Cao, B. Ren, M. M. Cheng and A. Borji, “Enhanced-alignment measure for sbinary foreground map evaluation,” arXiv preprint arXiv:1805.10421, 2018.10.24963/ijcai.2018/97 ↗
30. Salient object detection: A benchmark10.1109/TIP.2015.2487833 ↗
31. How to evaluate foreground maps