Two-Tier Submodel Partition Framework for Enhancing UAV Swarm Robustness in Forest Fire Detection
Bibliographic record
Abstract
The deployment of Unmanned Aerial Vehicle (UAV) swarm for Forest Fire Detection (FFD) missions presents unique challenges, e.g., the early forest fires are difficult to identify due to environment diversity and feature complexity, especially when some UAVs could be destroyed in harsh environments. To address these challenges, UAV swarm-based FFD missions can leverage advanced deep learning techniques, where online model updates, robustness, and communication overhead control become crucial for ensuring the effectiveness and adaptability of these missions. In this paper, we propose a Two-tier Submodel Partition Framework (TSPF) to enhance the robustness of UAV swarm conducting FFD missions. TSPF utilizes online model updates to adapt to diverse mission environments, thus strengthening the generalization capability of the model. In addition, a graph coloring method, an intragroup backup mechanism, and a Dynamic Server Selection (DSS) mechanism for the grouping are employed to enhance the robustness of FFD missions when some UAVs are destroyed, hence maintaining the high performance of FFD missions in harsh environments. Moreover, TSPF enables submodel updates by aggregating the parameters of selected layers within/between UAV groups, thereby effectively reducing the model parameter uploads (communication overhead) in model training. Experimental evaluations demonstrate that our proposed TSPF significantly improves the detection accuracy of forest fires, enhances the robustness of FFD missions against the destruction of some UAVs, and reduces the communication overhead in FFD missions.
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 | 0.002 |
| 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 | 0.001 |
| 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 | 0.000 |
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".