Edge computing-based computer vision and deep transfer learning for high-throughput assessment of Aspergillus flavus infection in crop seeds
Bibliographic record
Abstract
Manual assessment of toxic fungal infection levels in crop seeds is important for developing antifungal-resistant cultivars, yet it has long been recognized as health-risking and inherently subjective. This study presents an edge computing-based computer vision approach for high-throughput on-site assessment and quantification of Aspergillus flavus infection in crop seeds. The edge computing-based computer vision approach, termed Edge CV, was developed using the Jetson Nano, embedded cameras, and deployed with the proposed Edge CV model to enable intelligent evaluation with constrained computing resources and GPU power. The Edge CV model: First, leveraging semantic segmentation in computer vision tasks to differentiate between A. flavus -infected and uninfected; Second, utilizing post-processing techniques to accurately separate connected peanut seeds while merging segments belonging to the same ones; Third, analyzing and quantifying infection indices, as well as results presentation. Finally, deep transfer learning was employed to validate the model’s transferability for other crop seeds. As a result, Edge CV inference showed agreement with manual measurements (R 2 = 0.901, RMSE = 0.07) and superior consistency, with only a 0.01% fluctuation compared to 4.2% for human assessments. Moreover, Edge CV demonstrated its transferability to other crop seeds, such as maize (R 2 = 0.968, RMSE = 0.13) and rice (R 2 = 0.949, RMSE = 0.26). These results underscore the potential of Edge CV as a transferable solution for assessing toxic fungal infections. The approach developed also offers valuable insights for enhancing proximal machine vision, improving the distinction of adjacent seeds, and enabling more accurate calculation of the infection index.
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) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| 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".