End-to-end deep learning framework for printed circuit board manufacturing defect classification
Bibliographic record
Abstract
We report a complete deep-learning framework using a single-step object detection model in order to quickly and accurately detect and classify the types of manufacturing defects present on Printed Circuit Board (PCBs). We describe the complete model architecture and compare with the current state-of-the-art using the same PCB defect dataset. These benchmark methods include the Faster Region Based Convolutional Neural Network (FRCNN) with ResNet50, RetinaNet, and You-Only-Look-Once (YOLO) for defect detection and identification. Results show that our method achieves a 98.1% mean average precision(mAP[IoU = 0.5]) on the test samples using low-resolution images. This is 3.2% better than the state-of-the-art using low-resolution images (YOLO V5m) and 1.4% better than the state-of-the-art using high-resolution images (FRCNN-ResNet FPN). While achieving better accuracies, our model also requires roughly 3× fewer model parameters (7.02M) compared with the state-of-the-art FRCNN-ResNet FPN (23.59M) and YOLO V5m (20.08M). In most cases, the major bottleneck of the PCB manufacturing chain is quality control, reliability testing and manual rework of defective PCBs. Based on the initial results, we firmly believe that implementing this model on a PCB manufacturing line could significantly increase the production yield and throughput, while dramatically reducing manufacturing costs.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| 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 | 0.003 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.003 |
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".