Using SegFormer for Effective Semantic Cell Segmentation for Fault Detection in Photovoltaic Arrays
Bibliographic record
Abstract
Photovoltaic (PV) industries are susceptible to manufacturing defects within their solar cells. To accurately evaluate the efficacy of solar PV modules, the identification of manufacturing defects is imperative. Conventional industrial defect inspections predominantly rely on highly skilled inspectors conducting manual defect assessments, leading to sporadic and subjective identification outcomes. Deep-learning-based fault detection in PV or solar cells has emerged as a primary research area due to its superior efficiency and applicability. Hence, this study introduces a SegFormer-based fault detection framework to automate the visual defect inspection process in PV modules, complete with defect pseudocolorization. The proposed SegFormer-based framework effectively classifies defects into five categories: crack defects, front grid defects, interconnect defects, contact corrosion defects, and bright disconnect. Moreover, a comparative analysis is performed between the SegFormer model and the state-of-the-art fault detection algorithms, such as Deeplab v3, UNET, Deeplab v3+, PAN, PSPNet, and feature pyramid network (FPN). The experimental results reveal that the proposed SegFormer-based framework achieves highly encouraging performance, with a pixelwise accuracy of 96.24%, a weightedF1-score of 96.22%, an unweightedF1-score of 81.96%, and a mean intersection over union of 56.54%, outperforming other existing methods.
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 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".