A Multimodal UAV-Based Pipeline for Precision Agriculture: Aerial Stress Detection with YOLO and High-Fidelity Disease Classification Using DeiT
Bibliographic record
Abstract
Precision agriculture requires scalable, accurate monitoring solutions to bridge the gap from macro-level field awareness down to leaf-level diagnostics. This paper presents a novel multimodal pipeline that integrates Unmanned Aerial Vehicle (UAV) imagery with advanced deep-learning models, achieving broad-spectrum stress detection and fine-grained disease classification. A CNN initially established a foundational accuracy but struggled to capture the subtle signatures of different diseases. Switching to Vision Transformers (ViTs) improved performance, yet computational overhead and data requirements posed challenges. Later, we moved to DeiT with extensive hyperparameter tuning and data augmentations that gave an astonishing 99.45% accuracy on the multi-class plant disease dataset. To complement the close-up acumen of DeiT, we used YOLO from an aerial view to rapidly identify stressed versus healthy crop regions from real-time UAV footage. This two-tiered approach, wherein YOLO is used for aerial scanning and DeiT for leaf-level diagnoses, offers an unprecedented level of precision with scalability. Finally, complex output is translated by an LLM into farmer-friendly advisories to ensure immediate actionable insights. Our integrated framework sets a new benchmark in UAV-driven precision agriculture by balancing model sophistication, computational feasibility, and end-user interpretability.
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 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 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".