U-YOLO: Improved YOLOv5 for Small Object Detection on UAV-Captured Images

Zhang, Guowei; Chen, Xingyu; Tan, Xiao; Zhang, Jiahao; Lan, Xuguang

doi:10.1007/978-981-99-2789-0_1

Cited by 3 publications

(2 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to enhance the precision of object detection, we strive to improve the widely used YOLOv7 algorithm, which is depicted in Figure 2 as its network structure. There are many improved networks based on the YOLO network, and some are improved by using a single module [37][38][39], some are embedded in the network based on the large model [40][41][42], and some add new structures to the original network [43]. At the same time, there are also many improvements in small object detection based on the YOLO series in different application scenarios [44][45][46][47].…”

Section: Improvement Of Network Structurementioning

confidence: 99%

A Small Object Detection Method for Drone-Captured Images Based on Improved YOLOv7

Zhao,

Shao,

Liu

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

Due to the broad usage and widespread popularity of drones, the demand for a more accurate object detection algorithm for images captured by drone platforms has become increasingly urgent. This article addresses this issue by first analyzing the unique characteristics of datasets related to drones. We then select the widely used YOLOv7 algorithm as the foundation and conduct a comprehensive analysis of its limitations, proposing a targeted solution. In order to enhance the network’s ability to extract features from small objects, we introduce non-strided convolution modules and integrate modules that utilize attention mechanism principles into the baseline network. Additionally, we improve the semantic information expression for small targets by optimizing the feature fusion process in the network. During training, we adopt the latest Lion optimizer and MPDIoU loss to further boost the overall performance of the network. The improved network achieves impressive results, with mAP50 scores of 56.8% and 94.6% on the VisDrone2019 and NWPU VHR-10 datasets, respectively, particularly in detecting small objects.

show abstract

Section: Improvement Of Network Structurementioning

confidence: 99%

A Small Object Detection Method for Drone-Captured Images Based on Improved YOLOv7

Zhao,

Shao,

Liu

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…This improved detection performance over previous approaches was demonstrated by metrics like 3.9% mAP and 2.0% AP50 on the VisDrone dataset. Tang et al [34] presented HIC-YOLOv5, an enhanced YOLOv5 model that employs a small object detection head for high-resolution feature maps, an involution block for channel information enhancement, and CBAM for feature importance. As a result, the proposed model achieved 36.95% mAP:0.5.…”

Section: Related Workmentioning

confidence: 99%

PVswin-YOLOv8s: UAV-Based Pedestrian and Vehicle Detection for Traffic Management in Smart Cities Using Improved YOLOv8

Tahir,

Long,

Zhang

et al. 2024

Drones

View full text Add to dashboard Cite

In smart cities, effective traffic congestion management hinges on adept pedestrian and vehicle detection. Unmanned Aerial Vehicles (UAVs) offer a solution with mobility, cost-effectiveness, and a wide field of view, and yet, optimizing recognition models is crucial to surmounting challenges posed by small and occluded objects. To address these issues, we utilize the YOLOv8s model and a Swin Transformer block and introduce the PVswin-YOLOv8s model for pedestrian and vehicle detection based on UAVs. Firstly, the backbone network of YOLOv8s incorporates the Swin Transformer model for global feature extraction for small object detection. Secondly, to address the challenge of missed detections, we opt to integrate the CBAM into the neck of the YOLOv8. Both the channel and the spatial attention modules are used in this addition because of how well they extract feature information flow across the network. Finally, we employ Soft-NMS to improve the accuracy of pedestrian and vehicle detection in occlusion situations. Soft-NMS increases performance and manages overlapped boundary boxes well. The proposed network reduced the fraction of small objects overlooked and enhanced model detection performance. Performance comparisons with different YOLO versions ( for example YOLOv3 extremely small, YOLOv5, YOLOv6, and YOLOv7), YOLOv8 variants (YOLOv8n, YOLOv8s, YOLOv8m, and YOLOv8l), and classical object detectors (Faster-RCNN, Cascade R-CNN, RetinaNet, and CenterNet) were used to validate the superiority of the proposed PVswin-YOLOv8s model. The efficiency of the PVswin-YOLOv8s model was confirmed by the experimental findings, which showed a 4.8% increase in average detection accuracy (mAP) compared to YOLOv8s on the VisDrone2019 dataset.

show abstract