Downsampling, which aims to improve computational efficiency by reducing the spatial resolution of feature maps, is a critical operation in neural networks. Many downsampling methods have been proposed to address the challenge of retaining feature map information. However, some detailed information is still lost, even though these methods can extract features with stronger semantics. In this paper, we propose a novel downsampling method which combines feature slicing and depthwise separable convolution for information-retaining downsampling. It slices the input feature map into multiple non-overlapping sub-feature maps by using indexes with a stride of two in the spatial dimension and applies depthwise separable convolution on each slice to extract feature information. To demonstrate the effectiveness of SliceSamp, we compare it with classical downsampling methods on image classification, object detection, and semantic segmentation tasks using several benchmark datasets, including ImageNet-1K, COCO, VOC, and ADE20K. Extensive experiments demonstrate that SliceSamp outperforms classical downsampling methods with consistent improvements in various computer vision tasks. The proposed SliceSamp shows advanced model performance with lower computational costs and memory requirements. By replacing the downsampling layers in different network architectures (including ResNet (Residual Network), YOLOv5, and Swin Transformer), SliceSamp brings different degrees of performance gains (+0.54~3.64%) compared to these baseline models. Additionally, SliceUpsamp enables high-resolution feature reconstruction and alignment during upsampling. SliceSamp and SliceUpsamp can be plug-and-play-integrated into existing neural network architectures. As a promising downsampling alternative to traditional methods, SliceSamp can also provide a reference for designing lightweight and high-performance model architectures in the future.

With the advancement of computer vision, artificial intelligence, and remote sensing technologies, deep learning algorithms are increasingly used in terrestrial, airborne, and spaceborne-based fire detection systems. The performance and generalization of these data-driven fire detection algorithms, however, are restricted by the limited number and source of fire detection datasets. A large-scale fire detection benchmark dataset covering complex and varied fire scenarios is urgently needed. This work constructs a 100,000-level Flame and Smoke Detection Dataset (FASDD) based on multi-source heterogeneous flame and smoke images. It holds rich variations in image size, resolution, illumination (day and night), scenario (indoor and outdoor), image range (far and near), viewing angle (top view and side view), platform (surveillance cameras, drones, and satellites), and data source (Internet, social media, and open-access fire datasets). To the best of our knowledge, FASDD is currently the most versatile and comprehensive dataset for fire detection. It provides a challenging benchmark to drive the continuous evolution of fire detection models. Additionally, we formulate a unified workflow for preprocessing, annotation, and quality control of fire samples. Out-of-the-box annotations are published in four different formats for training deep learning models. Extensive performance evaluations based on classical methods show that most of the object detection models trained on FASDD can achieve satisfactory fire detection results, and especially YOLOv5x achieves nearly 80 % mAP@0.5 accuracy on heterogeneous images spanning two domains of computer vision and remote sensing. And the application in wildfire location demonstrates that deep learning models trained on our dataset can be used in recognizing and monitoring forest fires. Deep learning models trained with FASDD can be simultaneously deployed on satellites, drones, and ground sensors, thus realizing collaborative fire observation and detection in a space-air-ground integrated network environment. The dataset is available from the Science Data Bank website at https://doi.org/10.57760/sciencedb.j00104.00103 (Wang et al., 2022a).

Early warning systems on edge devices such as satellites and unmanned aerial vehicles (UAVs) are essential for effective forest fire prevention. Edge Intelligence (EI) enables deploying deep learning models on edge devices; however, traditional convolutional neural networks (CNNs)/Transformer-based models struggle to balance local-global context integration and computational efficiency in such constrained environments. To address these challenges, this paper proposes HybriDet, a novel hybrid-architecture neural network for wildfire detection. This architecture integrates the strengths of both CNNs and Transformers to effectively capture both local and global contextual information. Furthermore, we introduce efficient attention mechanisms—Windowed Attention and Coordinate-Spatial (CS) Attention—to simultaneously enhance channel-wise and spatial-wise features in high-resolution imagery, enabling long-range dependency modeling and discriminative feature extraction. Additionally, to optimize deployment efficiency, we also apply model pruning techniques to improve generalization performance and inference speed. Extensive experimental evaluations demonstrate that HybriDet achieves superior feature extraction capabilities while maintaining high computational efficiency. The optimized lightweight variant of HybriDet has a compact model size of merely 6.45 M parameters, facilitating seamless deployment on resource-constrained edge devices. Comparative evaluations on the FASDD-UAV, FASDD-RS, and VOC datasets demonstrate that HybriDet achieves superior performance over state-of-the-art models, particularly in processing highly heterogeneous remote sensing (RS) imagery. When benchmarked against YOLOv8, HybriDet demonstrates a 6.4% enhancement in mAP50 on the FASDD-RS dataset while maintaining comparable computational complexity. Meanwhile, on the VOC dataset and the FASDD-UAV dataset, our model improved by 3.6% and 0.2%, respectively, compared to the baseline model YOLOv8. These advancements highlight HybriDet’s theoretical significance as a novel hybrid EI framework for wildfire detection, with practical implications for disaster emergency response, socioeconomic security, and ecological conservation.

Deep learning methods driven by in situ video and remote sensing images have been used in fire detection. The performance and generalization of fire detection models, however, are restricted by the limited number and modality of fire detection training datasets. A large-scale fire detection benchmark dataset covering complex and varied fire scenarios is urgently needed. This work constructs a 100,000-level Flame and Smoke Detection Dataset (FASDD) based on multi-source heterogeneous flame and smoke images. To the best of our knowledge, FASDD is currently the most versatile and comprehensive dataset for fire detection. It provides a challenging benchmark to drive the continuous evolution of fire detection models. Additionally, we formulate a unified workflow for preprocessing, annotation and quality control of fire samples. Meanwhile, out-of-the-box annotations are published in four different formats for training deep learning models. Deep learning models trained on FASDD demonstrate the potential value and challenges of our dataset in fire detection and localization. Extensive performance evaluations based on classical methods show that most of the models trained on FASDD can achieve satisfactory fire detection results, and especially YOLOv5x achieves nearly 80 % mAP@0.5 accuracy on heterogeneous images spanning two domains of computer vision and remote sensing. And the application in wildfire location demonstrates that deep learning models trained on our dataset can be used in recognizing and monitoring forest fires. It can be deployed simultaneously on watchtowers, drones and optical satellites to build a satellite-ground cooperative observation network, which can provide an important reference for large-scale fire suppression, victim escape, firefighter rescue and government decision-making. The dataset is available from the Science Data Bank website at https://doi.org/10.57760/sciencedb.j00104.00103 (Wang et al., 2022).