A Lightweight Feature-Fusion and Small-Target Enhancement Network (LFE-YOLO), a lightweight detector that coordinates partial-channel feature extraction, efficient cross-scale fusion, high-resolution prediction, background-interference suppression, and stable tiny-box regression within a unified architecture is proposed.
Abstract
Detecting small unmanned aerial vehicles (UAVs) in ground-to-air imagery is challenging because their weak visual cues must be preserved without imposing excessive computation on resource-constrained platforms. To address the unresolved trade-off between tiny-target representation and deployment efficiency, we propose a Lightweight Feature-Fusion and Small-Target Enhancement Network (LFE-YOLO), a lightweight detector that coordinates partial-channel feature extraction, efficient cross-scale fusion, high-resolution prediction, background-interference suppression, and stable tiny-box regression within a unified architecture. Specifically, C2fFaster and GSConv reduce redundant computation while maintaining multi-scale feature propagation; a P2 high-resolution detection branch and Efficient Multi-scale Attention preserve fine spatial cues and suppress background interference; and Normalized Wasserstein Distance complements Complete Intersection over Union to improve tiny-box localization stability. We also construct Det-UAV by integrating newly collected multi-platform and multi-scene UAV imagery with existing data using scene- and sequence-independent partitioning and duplicate control. Experiments on Det-UAV show that LFE-YOLO improves detection accuracy while reducing parameters and computation relative to YOLOv8s. Zero-shot evaluation on the public DUT Anti-UAV dataset further indicates favorable transferability to an unseen data distribution. TensorRT 8.2.1 deployment experiments on NVIDIA Jetson TX2 show that, under the same evaluation settings, LFE-YOLO achieves higher detection accuracy and inference throughput, lower latency, and a smaller engine size than the comparable-scale YOLOv8s and YOLO11s models. These results support a practical accuracy–efficiency balance for small-UAV detection under constrained resources.
In complex low-altitude environments, illegally operated UAVs are typically small and can easily be confused with background clutter, leading to reduced detection stability. To address this challenge, we propose UAV-SGDFNet, an efficient small UAV detector. The model enhances robustness by jointly optimizing feature representation and multi-scale fusion. In the feature extraction phase, we introduce a spatial-channel collaborative attention mechanism to strengthen discriminative features for the target and suppress background interference. In the feature fusion phase, we design a structurally guided dual-level fusion module that enables structured cross-level interaction through the cooperation of a guidance path and a fidelity path, mitigating semantic dilution and improving the localization of small targets. Experimental results on a complex-scene dataset show that the proposed method outperforms several mainstream detectors, achieving an mAP@0.5 of 0.748 and increasing Recall to 0.692, while maintaining high inference efficiency, indicating a favorable accuracy-efficiency trade-off. Ablation studies and visual analyses further confirm the effectiveness of the proposed components for stable small-target detection and indicate its potential for real-time deployment.
Yang Yu, Kui-Xian Cai, Rui Fan et al.· Autonomous Intelligent Syste...· 0 citations
HD-YOLO improves small-object detection with a compact parameter footprint, while direct hardware benchmarks remain necessary to establish deployment efficiency.
Maosheng Sun, Jing Ding, Yang Zhang et al.· Applied Sciences· 0 citations
Unmanned aerial vehicle (UAV) imagery is widely used in urban monitoring, public security, and disaster assessment. However, object detection in UAV scenes faces multiple challenges, including a high proportion of small objects, severe occlusion in crowded areas, complex background textures, and image degradations such as haze, which often cause generic detectors to suffer from missed detections, false alarms, and unstable localization. To address these issues, we propose a lightweight multi-scale enhanced detection framework tailored for complex UAV scenarios. Built upon a MobileNet backbone, the proposed framework introduces a multi-scale enhancement module that constructs a feature pyramid and incorporates a scale-adaptive fusion mechanism to dynamically reweight the contributions of features from different scales. In addition, a fine-grained detail enhancement branch is deployed at high-resolution levels to strengthen edge and texture cues, while a context compensation module is designed to alleviate local uncertainty under dense occlusion and low-contrast conditions, thereby improving small-object separability and localization stability. Experimental results demonstrate that our method achieves strong performance on both VisDrone-DET and HazyDet, reaching mAP@0.5 = 0.312 and mAP@0.5:0.9 = 0.167 on VisDrone-DET, and mAP@0.5 = 0.719 and mAP@0.5:0.9 = 0.483 on HazyDet. The proposed method also shows favorable efficiency on an RTX 4060 Ti desktop GPU, indicating its real-time inference potential under the reported desktop hardware setting.
Xuehua Tao, Ji-Wei Sun· Engineering Research Express· 0 citations
Although unmanned aerial vehicle (UAV) platforms provide flexible deployment and wide coverage for aerial surveillance, high-altitude dynamic viewpoints usually involve small-scale targets, dense object distributions, and frequent occlusions, which bring significant challenges to detection and tracking. To address these issues, this paper proposes an efficient multi-level visual perception framework for UAV-based joint object detection and tracking. At the detection level, the proposed method is built upon YOLOv8n, removes the original P5 branch for large-object detection, and introduces a high-resolution P2 detection head for small objects. The backbone is further redesigned in a lightweight manner, and a multi-level fusion strategy is introduced to strengthen the perception ability for small targets. At the tracking level, the state modeling in ByteTrack is improved by refining the Kalman filtering process, which enhances trajectory continuity and identity consistency under dynamic viewpoints. Experimental results on the aerial surveillance test site indicate that the proposed method significantly improves detection and tracking performance while achieving a favorable trade-off between accuracy and computational efficiency, providing a practical reference for subsequent applications in UAV-based target monitoring.
