A Hybrid Deep Learning Approach for Early Detection of Wildfires Using Transformer Models and Feature Fusion Techniques
Open Access
- 29 September 2026
- journal article
- research article
- Published by MDPI AG in Electronics
- Vol. 15 (19) , 4470
Abstract
Natural disasters, particularly wildfires, pose a serious threat to ecosystems, human lives, and infrastructure, making rapid and accurate fire detection essential for effective disaster management. Despite recent advances in deep learning-based wildfire detection, conventional CNN-based approaches are inherently limited in modeling long-range spatial dependencies, while relying on a single deep architecture may provide an incomplete representation of complex and multi-scale fire patterns. Furthermore, high-dimensional deep features may contain redundant information, increasing computational complexity and limiting the efficiency of real-time monitoring systems. Therefore, there remains a need for an efficient framework that can exploit complementary global representations from multiple transformer architectures while retaining only the most discriminative features. To investigate this application-specific problem, this study evaluates a hybrid framework that integrates three heterogeneous Vision Transformer architectures—DeiT3, MaxViT, and Swin—with feature-level fusion, mRMR-based feature selection, and SVM classification for binary wildfire image classification. The publicly available FlameVision dataset, consisting of aerial images categorized into fire and non-fire classes, was used for evaluation. Deep feature representations extracted from the three transformer architectures were systematically fused to exploit their complementary characteristics. Minimum Redundancy Maximum Relevance (mRMR) feature selection was subsequently employed to reduce feature redundancy and retain the most discriminative information, followed by classification using Support Vector Machines (SVM). The experimental results demonstrated an overall classification accuracy of up to 100% on the internal hold-out test partition and in five-fold cross-validation within the FlameVision dataset. Because both evaluation procedures were conducted using images originating from the same source dataset, these results should be interpreted as within-dataset classification performance rather than evidence of external generalization. Validation on independently collected wildfire datasets is therefore required to determine the robustness of the proposed framework under different geographical, sensor, environmental, and acquisition conditions. The proposed framework demonstrates that complementary transformer-derived representations combined with feature selection can provide highly discriminative wildfire representations while substantially reducing the number of features required for classification. These findings demonstrate that the proposed approach can retain high classification performance using a substantially more compact feature representation. However, the resulting reduction in feature dimensionality should not be interpreted as evidence of improved runtime efficiency, since computational time before and after feature selection was not experimentally measured. However, real-time performance is not claimed in the present study because FPS, end-to-end latency, runtime memory consumption, and deployment on continuous UAV or satellite streams were not experimentally evaluated. Such operational evaluation is left for future work.Keywords
This publication has 53 references indexed in Scilit:
- Development and Application of a Sentinel-2 Satellite Imagery Dataset for Deep-Learning Driven Forest Wildfire DetectionSN Computer Science, 2025
- An advanced multi-source data fusion method utilizing deep learning techniques for fire detectionEngineering Applications of Artificial Intelligence, 2024
- A systematic review of trustworthy artificial intelligence applications in natural disastersComputers and Electrical Engineering, 2024
- Wildfire Impact Analysis and Spread Dynamics Estimation on Satellite Images Using Deep LearningJournal of the Indian Society of Remote Sensing, 2024
- Wildfire response of forest species from multispectral LiDAR data. A deep learning approach with synthetic dataEcological Informatics, 2024
- Fed-mRMR: A lossless federated feature selection methodInformation Sciences, 2024
- Creation of wildfire susceptibility maps in the Mediterranean Region (Turkey) using convolutional neural networks and multilayer perceptron techniquesForest Ecology and Management, 2023
- The impact of forest fires on floods and erosion: Marmaris, TurkeyEnvironment, Development and Sustainability, 2022
- A new brain tumor diagnostic model: Selection of textural feature extraction algorithms and convolution neural network features with optimization algorithmsComputers in Biology and Medicine, 2022
- Detection of lung cancer on chest CT images using minimum redundancy maximum relevance feature selection method with convolutional neural networksBiocybernetics and Biomedical Engineering, 2020