Due to the limited computing resources and storage capacity of edge detection devices, efficient detection algorithms are typically required to meet real-time and accuracy requirements. Existing detectors often require a large number of parameters and high computational power to improve accuracy, which reduces detection speed and performance on low-power devices. To reduce computational load and enhance detection performance on edge devices, we propose a lightweight drone target detection algorithm, LD-YOLOv10. Firstly, we design a novel lightweight feature extraction structure called RGELAN, which utilizes re-parameterized convolutions and the newly designed Conv-Tiny as the computational structure to reduce the computational burden of feature extraction. The AIFI module was introduced, utilizing its multi-head attention mechanism to enhance the expression of semantic information. We construct the DR-PAN Neck structure, which obtains weak features of small targets with minimal computational load. Wise-IoU and EIoU are combined as new bounding box regression loss functions to adjust the competition between anchor boxes of different quality and the sensitivity of anchor box aspect ratios, providing a more intelligent gradient allocation strategy. Extensive experiments on the VisdroneDET-2021 and UAVDT datasets show that LD-YOLOv10 reduces the number of parameters by 62.4% while achieving a slight increase in accuracy and has a faster detection speed compared to other lightweight algorithms. When deployed on the low-power NVIDIA Jetson Orin Nano device, LD-YOLOv10 achieves a detection speed of 25 FPS.

In the realm of traffic sign detection, challenges arise due to the small size of objects, complex scenes, varying scales of signs, and dispersed objects. To address these problems, this paper proposes a small object detection algorithm, YOLOv8s-DDA, for traffic signs based on an improved YOLOv8s. Specifically, the C2f-DWR-DRB module is introduced, which utilizes an efficient two-step method to capture multi-scale contextual information and employs a dilated re-parameterization block to enhance feature extraction quality while maintaining computational efficiency. The neck network is improved by incorporating ideas from ASF-YOLO, enabling the fusion of multi-scale object features and significantly boosting small object detection capabilities. Finally, the original IoU is replaced with Wise-IoU to further improve detection accuracy. On the TT100K dataset, the YOLOv8s-DDA algorithm achieves mAP@0.5 of 87.2%, mAP@0.5:0.95 of 68.3%, precision of 85.2%, and recall of 80.0%, with a 5.4% reduction in parameter count. The effectiveness of this algorithm is also validated on the publicly available Chinese traffic sign detection dataset, CCTSDB2021.

Object detection in low-light environments has widespread applications, but existing low-light detection algorithms often suffer from issues such as the loss of scarce low-level features in low-light images, poor performance in detecting multi-scale objects, and reduced accuracy in locating objects with blurred edges or occlusions, resulting in suboptimal performance in low-light scenarios. To address these challenges, we propose a novel feature fusion structure, SRB-FPN, and optimize the regression loss function, integrating it with the YOLOv8 model to create a detector, LLD-YOLO, specifically designed for low-light conditions. To emphasize the key unit information in feature maps, while dynamically adjusting the fusion ratio between shallow and deep features based on their importance, we introduce the SDFBF module to preserve crucial low-level semantics.To enhance the diversity of receptive fields during training, enabling the capture of multi-scale features, while avoiding additional inference overhead, we incorporate the DBB-c2f module.To make the model focus more on challenging samples in low-light conditions and more accurately capture the differences in bounding box boundaries, we combine Focaler IoU and MPDIoU, and propose the Focal MPDIoU Loss. Finally, the proposed model achieves a mAP50 of 70.0% on the ExDark dataset, which is a 2.7% improvement over the baseline. Extensive experiments on the ExDark, NOD, and RTTS datasets demonstrate that LLD-YOLO exhibits exceptional detection performance in low-light environments and strong adaptability in foggy weather scenarios.

Object detection in remote sensing images is of significant research value in fields such as environmental monitoring and urban planning. However, the large variation in object sizes, along with challenges such as small and densely packed objects, makes this task particularly challenging. To address these issues, we propose an algorithm for multi-scale feature extraction in remote sensing image detection using dilated residuals (ADS-YOLO). Firstly, to address the challenges of scale variation and small target size, the Dilation-wise Residual (DWR) design is employed to form the C2f_DWR module, which restructures the bottleneck structure within the C2f segment to facilitate the extraction and fusion of multiscale contextual information, thus reducing the difficulty associated with target scale variation. Secondly, inspired by the Adown subsampling convolution module from YOLOv9, we use it to replace the convolutions in the Backbone, enabling the model to capture finer image details at higher levels, while maintaining accuracy and reducing computational load. Lastly, to address the issue of dense targets, we design the Soft-NMS-ShapeIoU module to improve the consistency of bounding boxes and target shapes, while also suppressing adjacent boxes. Experimental results demonstrate that, on the publicly available remote sensing image datasets DIOR, RSOD, and NWPU VHR-10, the proposed ADS-YOLO model outperforms other state-of-the-art methods by a significant margin.

RGB-T object detection, by fusing complementary information from visible and infrared images, has wide applications in fields such as military reconnaissance, autonomous driving, and intelligent surveillance. Cross-modal feature fusion is a key component of this task. However, current approaches struggle with effective spatial-channel feature extraction and adaptive cross-modal calibration. By failing to fully exploit inter-modal complementarities, these methods often suppress key semantics or retain low-level noise, ultimately degrading the quality of the fused representation. To address these issues, this paper proposes a fusion network, DReAM-Fusion, driven by a dual-modal recalibrated feature aggregation module (DReAM). In DReAM, a channel emphasis convolution is designed to suppress redundant channels and reinforce key semantic information, while a channel-spatial collaborative attention mechanism is introduced to weight features across two dimensions, thereby enhancing the response of core information. Additionally, a shared group normalization detection head is designed, utilizing a shared convolution structure to reduce the feature distribution discrepancy between different detection layers. A learnable channel scaling layer is also incorporated, assigning trainable scaling factors to each channel, thus alleviating gradient instability issues. Lastly, we design a reparameterized multi-scale convolution (RMC). By incorporating dilated reparameterized convolution, RMC enhances scale adaptability while maintaining a constant computational cost through structural reparameterization. Experimental results show that DReAM-Fusion achieves mAP ₅₀ of 87.7% and mAP of 60.8% on the M3FD dataset. Compared to the current state-of-the-art model, cross-modality fusion transformer (CFT), the proposed method reduces the parameter count by 27.59 M (62%) while still improving detection performance by 0.6% and 5.4% points, respectively. Furthermore, experiments on the KAIST and VEDAI datasets further validate the comprehensive advantages of the proposed method in terms of detection accuracy and computational efficiency.