An Automatic Rice Plant Disease Classification Using Hierarchical Vision Transformers with Spatial Reduction Attention Mechanism
Bibliographic record
Abstract
Rice is a staple crop for world food security, but its yield is susceptible to several diseases.Timely recognition of rice leaf diseases (RLD) is critical to decrease yield losses.Recent studies offered improved solutions by harnessing deep networks to precisely diagnose and categorize RLDs.Conversely, due to extreme scale deviations, data redundancy, and the high-resolution multi-spectral nature of leaf images, the conventional deep-learning classifiers underperform in detecting rice diseases.Also, training deep learning models involves major challenges such as class imbalance, overfitting, and vanishing gradient problems.Nowadays, vision transformers are infiltrated into the field of image processing, in which the self-attention unit is employed to learn local and distant correlations among the pixels in an image.However, the processing and storage overheads of analyzing image patches are very high.In this context, we propose a new vision transformer-based RLD classifier, called Faster Hierarchical Vision Transformer (FHViT) which employs a Spatial Reduction Attention Mechanism (SRAM) to speed up the classification process.The SRAM module enables the transformer to estimate the significance of each pixel in image patches and optimize their effect on the result.We evaluate our model on an open-access Ade F. Rice Leaf Diseases (AFRLD) database and relate its performance with other advanced models in terms of performance indicators.Our model delivers 99.2% detection accuracy, 99.3% precision, 99.2% sensitivity, 99.1% specificity, 99.0% recall, and 99.0%F1 measure.The extensive experimentations demonstrate that the FHViT realizes a viable solution for RLD diagnosis.
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 | 0.000 |
| 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 | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".