Unmanned aerial vehicle (UAV) targets are typically small in size, occupy only a limited pixel area, and are often located in complex environments. Existing models, however, tend to overlook smaller targets in complex backgrounds, making it easy to miss important information and resulting in missing targets. This paper proposes an innovative UAV detection method called BRA-YOLOv10. Firstly, Bi-Level Routing Attention (BRA) is used during the feature extraction stage to effectively reduce background interference. By focusing on the target’s key information, the model optimizes overall detection performance. Secondly, based on this improvement, adding a high-resolution small target detection layer (STDL) enhances the model’s detection accuracy by providing more detailed target information. Then, the SimCSPSPPF module is incorporated to further improve the model’s performance, particularly with regard to small targets on complex backgrounds, to effectively retain low-level feature information. Finally, training and testing are conducted on the SIDD dataset, and its generalization performance is shown on a variety of datasets (TIB-Net, Det-Fly, and UAVfly). Experimental results show that BRA-YOLOv10 has superior performance to existing models on 98.9% precision, 92.3% recall, and 96.5% mAP50, respectively, representing an improvement of 2.5%, 2.9%, and 1.7% over YOLOv10. For targets with target sizes between $10\times 10$ and $5\times 5$, the detection rate reached 84.97%, an increase of 16.76% compared with YOLOv10.

Unmanned Aerial Vehicle (UAV) object detection has witnessed rapid progress in recent years. However, its heavy reliance on labeled data and the assumption of consistent data distributions between training and deployment domains limit its generalization ability, leading to significant performance degradation under domain shifts. To address this challenge arising from substantial discrepancies in feature distributions across UAV images captured under diverse conditions, we propose a novel framework: Adaptive Teacher with Weighted Multi-layer Feature Alignment (WMFA-AT) for cross-domain UAV object detection. WMFA-AT adopts a teacher–student mutual learning paradigm, integrating domain adversarial learning with weighted multi-layer feature alignment and strong-weak data augmentation to effectively mitigate domain discrepancies. Specifically, the student model performs adversarial alignment using multiple domain discriminators applied to different feature layers, where layer-wise transferability is quantitatively estimated and used to adaptively weight the alignment process. This strategy ensures that features from the source and target domains are aligned in a distribution-aware manner. Meanwhile, the teacher model benefits from the student model via mutual learning, incorporating knowledge from both source and target domains while avoiding overfitting to the source. To comprehensively evaluate the proposed approach, we construct four challenging cross-domain UAV object detection benchmarks covering cross-time, cross-camera, cross-view, and cross-weather scenarios. Experimental results demonstrate that WMFA-AT consistently improves detection accuracy across diverse domain shifts, highlighting its robustness, generalization capability, and practical applicability in real-world UAV deployment settings.

Recent years have witnessed the emergence of numerous U-shaped deep learning segmentation methods for infrared small-target detection (IRSTD). However, increasingly complex models still suffer from false and missed detections in challenging scenarios with cluttered backgrounds and weak targets while incurring escalating computational costs. To address these limitations, this paper proposes MCC-Net, a novel and efficient IRSTD framework that achieves superior detection performance with significantly reduced computational complexity. First, we integrate Magnitude-Aware Linear Attention (MALA) and Conditionally Parameterized Convolutions (CondConv) to replace conventional attention mechanisms in skip connections and standard convolutions, respectively, endowing the model with spatial contextual modeling and enhanced feature extraction capabilities at minimal computational overhead. Second, we design an innovative Conditional Cross-Channel Fusion (CondCCF) module that establishes a complementary spatial-channel dual-attention mechanism with MALA, enabling efficient multi-scale feature fusion. Extensive comparative and ablation experiments conducted on three public benchmarks—SIRST-v1, NUDT-SIRST, and IRSTD-1K—demonstrate that MCC-Net achieves state-of-the-art mIoU scores of 77.98%, 95.43%, and 70.46%, respectively, surpassing state-of-the-art methods by 1.07%, 1.95%, and 0.95%. MCC-Net also outperforms existing approaches across multiple evaluation metrics while maintaining substantially lower computational complexity.