Articles
An Integrated Lightweight UAV Vision and Relation-Aware Language Intelligence Framework for Early Wildfire Smoke Detection and Emergency Knowledge
Downloads
Keywords
wildfire smoke detection
UAV intelligence
lightweight deep learning
relation extraction
Article Usage Statistics
Live Metrics
Abstract Views
212
File Views
80
PDF
0
HTML
0
Other
80
Total
292
Article Usage Chart
Abstract
File Views
PDF
HTML
Other
Total
Abstract
Background: Early wildfire detection remains one of the most urgent technical and environmental challenges in disaster management. Recent advances in unmanned aerial vehicles, lightweight smoke detection networks, transfer learning, and autonomous sensing have substantially improved the feasibility of real-time aerial fire surveillance under resource-constrained conditions (Chen et al., 2023; Yang et al., 2023; Hossain & Zhang, 2023; Abdelfattah et al., 2025). At the same time, emergency response ecosystems increasingly rely on text-rich operational data including incident logs, drone reports, sensor annotations, and situational summaries. Progress in relation extraction, named entity recognition, transformer architectures, joint extraction methods, and knowledge-aware graph learning has created new opportunities to transform such unstructured textual data into actionable emergency knowledge (Han et al., 2019a; Devlin et al., 2019; Wei et al., 2020; Xu et al., 2020; Wang et al., 2021). However, these two streams are still usually studied in isolation.
Objective: This article develops a unified conceptual research framework that integrates lightweight UAV-based wildfire smoke detection with relation-aware language intelligence for structured emergency knowledge extraction.
Methods: Using only the provided references, the study conducts an integrative, architecture-oriented synthesis. The framework combines UAV autonomy, lightweight smoke segmentation and detection, multimodal event reporting, open information extraction, neural relation extraction, joint entity-relation extraction, transformer-based contextual modeling, and graph-based emergency knowledge construction.
Results: The analysis indicates that aerial smoke detection systems become significantly more operationally valuable when coupled with language models that extract entities, relations, event states, causal descriptions, and response dependencies from drone-generated and command-level text. Lightweight visual detection alone supports early warning, but integrated language intelligence enables prioritization, coordination, and structured operational understanding. The study also finds that relation extraction architectures based on joint extraction, attention, cascade tagging, token-pair linking, and adaptive thresholding are particularly suitable for emergency documentation streams characterized by sparse, evolving, and long-tailed relational patterns (Zheng et al., 2017; Wei et al., 2020; Qu et al., 2020).
Conclusion: A next-generation wildfire intelligence system should not treat aerial smoke recognition and emergency text understanding as separate modules. It should function as a unified sensing-to-knowledge pipeline in which UAV imagery supplies early visual evidence and language intelligence transforms fragmented operational text into structured
situational awareness. This integrated approach offers a strong conceptual basis for trustworthy, scalable, and response-oriented wildfire monitoring systems.
Downloads
Download data is not yet available.
References
1. Abdelfattah, R., Abdelfatah, K., Fouda, M. M., et al. (2025). Bayesian optimization-aided hybrid deep learning model for lightweight UAV-based smoke detection. IEEE Internet of Things Journal, 12(16), 33506–33519.
2. Alsalem, A., & Zohdy, M. (2024). Wheat field fire smoke detection from UAV images using CNN-CBAM. In 2024 2nd International Conference on Artificial Intelligence, Blockchain, and Internet of Things (AIBThings) (pp. 1–8).
3. Bhar, A., & Sayadi, M. (2024). On designing a configurable UAV autopilot for unmanned quadrotors. Frontiers in Neurorobotics, 18, 16.
4. Chen, G., Cheng, R., Lin, X., et al. (2023). LMDFS: A lightweight model for detecting forest fire smoke in UAV images based on YOLOv7. Remote Sensing, 15(15), 23.
5. Devlin, J., Chang, M. W., Lee, K., et al. (2019). BERT: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Vol. 1, pp. 4171–4186).
6. Divazi, A., Askari, R., & Roohi, E. (2025). Experimental and numerical investigation on the spraying performance of an agricultural unmanned aerial vehicle. Aerospace Science and Technology, 110083.
7. dos Santos, C. N., & Guimarães, V. (2015). Deep convolutional neural networks for relation classification. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing (pp. 1635–1640).
8. Etzioni, O., Banko, M., Soderland, S., et al. (2005). Open information extraction from the web. Communications of the ACM, 48(12), 68–74.
9. Gardent, C., Shimorina, A., Narayan, S., et al. (2017). WebNLG challenge: Generating text from RDF data. In Proceedings of the 10th International Conference on Natural Language Generation (pp. 124–133).
10. Guo, Z., Zhang, Y., Lu, W., et al. (2019). Knowledge-aware graph networks with label smoothness regularization for relation extraction. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (pp. 1920–1930).
11. Han, X., Yu, P., Liu, Z., et al. (2019a). A survey on neural relation extraction. IEEE Transactions on Knowledge and Data Engineering, 32(1), 5–29.
12. Han, X., Yu, P., Liu, Z., et al. (2019b). GraphRel: Modeling text as relational graphs for joint entity and relation extraction. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (pp. 1409–1418).
13. Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780.
14. Hossain, F. M. A., & Zhang, Y. (2023). MsFireD-Net: A lightweight and efficient convolutional neural network for flame and smoke segmentation. Journal of Automation and Intelligence, 2(3), 130–138.
15. Li, Q., & Ji, H. (2014). Incremental joint extraction of entity mentions and relations. In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Vol. 1, Issue 1, pp. 402–412).
16. Liu, X., & Zhang, L. (2017). Neural relation extraction with multi-lingual attention. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing (pp. 2660–2669).
17. Miwa, M., & Bansal, M. (2016). End-to-end relation extraction using LSTMs on sequences and tree structures. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Vol. 1, pp. 1105–1116).
18. Miwa, M., & Sasaki, Y. (2014). Modeling joint entity and relation extraction with table representation. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) (pp. 1858–1869).
19. Qu, Y., Chen, X., & Ren, X. (2020). ATLOP: Adaptive thresholding for long-tailed relation extraction. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) (pp. 5284–5293).
20. Raczok, T., Ivens, S. N., Seidel, L., et al. (2025). Wildfire detection and monitoring: A drone-based approach and comparative analysis. Proceedings of the International ISCRAM Conference.
21. Ragbir, P., Kaduwela, A., Lan, X., et al. (2024). A control-theoretic spatio-temporal model for wildfire smoke propagation using UAV-based air pollutant measurements. Drones, 8(5).
22. Silva, J., Sousa, D., Vaz, P., et al. (2024). Application of unmanned aerial vehicles for autonomous fire detection. In International Conference on Disruptive Technologies, Tech Ethics and Artificial Intelligence. Springer, Cham.
23. Soares, L. B., FitzGerald, N., Ling, J., et al. (2019). Matching the blanks: Distributional similarity for relation learning. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (pp. 2895–2905).
24. Vaswani, A., et al. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 30.
25. Wang, S., Han, X., & Zhao, J. (2021). A survey on deep learning for named entity recognition. IEEE Transactions on Knowledge and Data Engineering.
26. Wei, X., & Su, J. (2020). TPLinker: Single-stage joint extraction of entities and relations through token pair linking. arXiv preprint arXiv:2010.13415.
27. Wei, Z., Su, D., Wang, Y., et al. (2020). CasRel: A novel cascade binary tagging framework for relation extraction. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (pp. 1950–1960).
28. Xu, Y., Feng, Y., Huang, S., et al. (2020). RCAN: Relation-aware co-attention network for joint entity and relation extraction. Knowledge-Based Systems, 199, 105912.
29. Yang, H., Wang, J., & Wang, J. (2023). Efficient detection of forest fire smoke in UAV aerial imagery based on an improved YOLOv5 model and transfer learning. Remote Sensing, 15(23).
30. Zelenko, D., Aone, C., & Richardella, A. (2003). Kernel methods for relation extraction. Journal of Machine Learning Research, 3, 1083–1106.
31. Zeng, D., Liu, K., Lai, S., et al. (2014). Relation classification via convolutional deep neural network. In Proceedings of COLING 2014, the 25th International Conference on Computational Linguistics: Technical Papers (pp. 2335–2344).
32. Zeng, X., Zeng, D., He, S., et al. (2018). Extracting relational facts by an end-to-end neural model with copy mechanism. In Proceedings of the 27th International Conference on Computational Linguistics (pp. 5068–5078).
33. Zhang, Y., Zhong, V., & Chen, D. (2019). Position-aware attention and supervised data improve slot filling. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Vol. 1, pp. 221–231).
34. Zheng, S., Wang, F., Bao, H., et al. (2017). Joint extraction of entities and relations based on a novel tagging scheme. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Vol. 1, pp. 1227–1236).