Small object detection in aerial imagery remains challenging due to sparse feature representation, limited spatial resolution, and complex background interference. Current deep learning approaches enhance detection performance through multi-scale feature fusion, leveraging convolutional operations to expand the receptive field or self-attention mechanisms for global context modeling. However, these methods primarily rely on spatial-domain features, while self-attention introduces high computational costs, and conventional fusion strategies (e.g., concatenation or addition) often result in weak feature correlation or boundary misalignment. To address these challenges, we propose a unified spatial-frequency modeling and multi-scale alignment fusion framework, termed USF-DETR, for small object detection. The framework comprises three key modules: the Spatial-Frequency Interaction Backbone (SFIB), the Dual Alignment and Balance Fusion FPN (DABF-FPN), and the Efficient Attention-AIFI (EA-AIFI). The SFIB integrates the Scharr operator for spatial edge and detail extraction and FFT/IFFT for capturing frequency-domain patterns, achieving a balanced fusion of global semantics and local details. The DABF-FPN employs bidirectional geometric alignment and adaptive attention to enhance the significance expression of the target area, suppress background noise, and improve feature asymmetry across scales. The EA-AIFI streamlines the Transformer attention mechanism by removing key-value interactions and encoding query relationships via linear projections, significantly boosting inference speed and contextual modeling. Experiments on the VisDrone and TinyPerson datasets demonstrate the effectiveness of USF-DETR, achieving improvements of 2.3% and 1.4% mAP over baselines, respectively, while balancing accuracy and computational efficiency. The framework outperforms state-of-the-art methods in small object detection.

Rotated object detection is a crucial task in aerial image analysis. To address challenges such as multi-directional object rotation, complex backgrounds with occlusions, and the trade-off between speed and accuracy in remote sensing images, this paper introduces a lightweight progressive fusion calibration network for rotated object detection (LPFC-RDet). The network comprises three main modules: the Retentive Meet Transformers (RMT) feature extraction block, the Progressive Fusion Calibration module (PFC), and the Shared Group Convolution Lightweight detection head (SGCL). The RMT feature extraction block integrates a retentive mechanism with global context modeling to learn rotation-insensitive features. The PFC module employs pixel-level, local-level, and global-level weights to calibrate features, enhancing feature extraction from occluded objects while suppressing background interference. The SGCL detection head uses decoupled detection tasks and shared group convolution layers to achieve parameter sharing and feature interaction, improving accuracy while maintaining a lightweight structure. Experimental results demonstrate that our method surpasses state-of-the-art detectors on three widely used remote sensing object datasets: HRSC2016, UCAS_AOD, and DOTA.

Infrared and visible image fusion-based object detection is critical for robust environmental perception under adverse conditions, yet existing methods still suffer from insufficient modeling of modality discrepancies and limited adaptivity in their fusion mechanisms. This work proposes an asymmetric spatial–frequency fusion network, AsyFusionNet. The network adopts an asymmetric dual-branch backbone that extends the RGB branch to P5 while truncating the infrared branch at P4, thereby better aligning with the physical characteristics of the two modalities, enhancing feature complementarity, and enabling fine-grained modeling of modality differences. On top of this backbone, a local–global attention fusion (LGAF) module is introduced to model local and global attention in parallel and reorganize them through lightweight convolutions, achieving joint spatial–channel selective enhancement. Modality-specific feature enhancement is further realized via a hierarchical attention module (HAM) in the RGB branch, which employs dynamic kernel selection to emphasize multi-level texture details, and a fourier spatial spectral modulation (FS2M) module in the infrared branch, which more effectively captures global thermal radiation patterns. Extensive experiments on the $M^3$FD and VEDAI datasets demonstrate that AsyFusionNet attains $86.3\%$ and $54.1\%$${\mathrm{mAP}}_{50}$, respectively, surpassing the baseline by $8.8$ and $6.4$ points (approximately $11.4\%$ and $13.4\%$ relative gains) while maintaining real-time inference speed.