Metal Structural Defect Detection Based-On Deep Learning and Grad-Cam
Bibliographic record
Abstract
It is crucial to guarantee the quality of the surface of metal products, with different inspection methods and technologies being recommended recently. Traditional techniques for manually inspecting items face several limitations and often find it challenging to ensure flawless results. Vision-based approaches for automatic examination of metal surfaces have emerged as powerful and effective methods for tackling various quality control challenges in the industrial sector. Therefore, in this study a surface defects detection for metal images using modified NasNetMobile and three other convolution neural network. Furthermore, we conducted a comparison study between modified NasNetMobile, MobileNetV2, InceptionV3, RensNet50. We have trained and tested our classification models using a public database of Northeastern University composed of 1800 images of defects. Evaluation phase showed that the the modified lightweight models including MobileNetV2 and NasNetMobile achieved good results with 99.7% of accuracy, while applying Grad-Cam algorithm demonstrate that our models can be easily utilized to efficiently inspect metal surface defects even if it is with background different of the images used in training. Testing results of our model on external images showed that the proposed study is capable of identifying and localizing the defected region on wind turbine surface and other metal panel types.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".