Transferable Learning Architecture for Scalable Visual Quality Inspection
Bibliographic record
Abstract
In recent years, convolutional neural networks (CNNs) have become a de facto standard in computer vision for object detection and recognition. At present, CNNs have been used in many application areas including the automation of industrial manufacturing processes. But using CNN in a real-time environment to track defects on products has many shortcomings like long training time, large data requirements, slow inference time, dynamic environment, and hardware dependency. This paper evaluates the state-of-the-art CNN architectures for object detection to address the mentioned challenges and provide the best possible solution. A set of pre-trained models has been trained on just 781 annotated images by applying transfer learning. Experimental results showed that Faster RCNN with VGG-16 backbone outperforms the other models in case of accuracy and mAP. But RetinaNet with an FPN backbone has the fastest inference time on multi-scaled defects. Paper also presents the deployment pipeline for inference on mobile devices to use in a real-time environment without any special hardware. In addition, an improved dataset of submersible pump impellers, based on the existing Kaggle dataset is introduced.
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 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.002 |
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".