LCA-Net: A Lightweight Network for Small Object Detection in Road Traffic Scenes

Small-Target Traffic Sign Detection Method Based on Multi-Path Feature Aggregation and Attention Enhancement

Detecting traffic signs in real-world roadway scenes remains a demanding task due to extensive category diversity, the prevalence of diminutive targets, and interference from cluttered surroundings. To overcome these obstacles, we present YOLO-PPA, a YOLOv11n-based detector strengthened through multi-path feature aggregation and attention-enhanced representation learning. First, a Parallelized Patch-Aware Attention (PPA) mechanism is embedded in place of the standard C3K2 block, simultaneously capturing fine-grained local textures and broad contextual semantics while adaptively amplifying informative spatial regions critical for small objects. Second, a high-resolution P2 detection head is appended to the feature pyramid, recovering fine spatial cues that would otherwise be attenuated across successive downsampling stages, and this design is particularly beneficial for recognizing signage occupying only a handful of pixels. Third, the Normalized Gaussian Wasserstein Distance (NWD) replaces the conventional CIoU metric as the regression loss, offering a smoother optimization landscape for tiny instances where even single-pixel displacements can destabilize standard IoU-based objectives. Evaluated on the TT100K benchmark, YOLO-PPA surpasses the YOLOv11n baseline by 2.1% in precision, 3.7% in recall, 4.3% in mAP@50, and 3.0% in mAP@50:95, confirming its effectiveness for small-scale traffic sign recognition in complex driving environments.

MSHC-YOLO: multiscale heterogeneous convolution for small object detection in traffic scenes

Small object detection in traffic monitoring suffers from a structural inefficiency in standard detectors: isotropic 3×3 convolutions treat all spatial directions uniformly, yet traffic objects exhibit strong anisotropic geometry—pedestrians are vertically elongated, vehicles are horizontally wide. We propose MSHC-YOLO, built on two complementary innovations. First, a Multi-Scale Heterogeneous Convolution (MSHC) module decomposes standard convolution into three parallel kernels with distinct shapes—square (3×3), horizontal strip (1×3), and vertical strip (3×1)—to capture direction-specific spatial structures at negligible parameter cost. Second, a P2-Guided Fine-Grained Detection Head extends the feature pyramid to stride 4, providing a 160×160 detection grid that resolves objects as small as 8×8 pixels. Critically, MSHC and the P2 head interact synergistically: heterogeneous kernels are most effective in shallow, highresolution layers where spatial detail is preserved. On VisDrone2019-DET, MSHC-YOLO achieves 37.5% mAP@0.5, outperforming YOLO11n by 4.2 points while adding only 0.21M parameters. Ablation reveals a super-linear gain— MSHC contributes +0.5 mAP in the standard P3 path but +2.0 mAP when combined with the P2 head, validating that the shallower the feature, the more kernel shape matters.

AG-Net: adaptive geometry-aware network for aerial oriented object detection

A robust small-object detection model for UAV aerial imagery under complex background clutter

To tackle the challenge of detecting small targets in UAV imagery, this paper proposes PF-DETR, an enhanced object detection model based on RT-DETR, designed to improve detection accuracy in complex scenes. The improvements are primarily reflected in the following aspects. First, a P2 detection head is added to extend the feature pyramid to finer scales, thereby enhancing the ability of shallow features to detect small targets. Second, a Pyramidal Hierarchical Frequency-Domain Fusion (PHF) module is introduced. By combining wavelet pooling with high- and low-frequency attention, the module effectively extracts and fuses multi-scale features, reduces information loss, and improves detection accuracy for small targets. Finally, the backbone network is restructured through the design of a lightweight BasicBlock_FasterNet_Rep module, which integrates FasterNet and RepVGG-style re-parameterization. This restructuring significantly reduces model complexity and parameters while strengthening multi-scale feature extraction. Experimental results on the VisDrone2019 dataset show that the improved PF-DETR achieves a notable performance boost: compared to the original model, mAP@0.5 increases by 5.4%, while the number of parameters is reduced by 25.0%. The computational cost increases by about 35.0%, but this comes with higher accuracy, resulting in a favorable balance between detection performance and model efficiency. Overall, these improvements enhance the model’s robustness and accuracy in detecting multi-scale and small targets in complex and cluttered scenes.

HASO-DETR: hybrid attention small object detection based on RT-DETR

The proposed framework features a redesigned cross-scale feature fusion module, CCFM-S2, which utilizes the SPD-Conv operator for information preserving downsampling and explicitly integrates high-resolution shallow features (S2 layer), thereby infusing indispensable spatial details into the feature hierarchy for small targets.

Context-gated dynamic perception for small-object detection in dense aerial scenes

A context-gated dynamic perception framework that treats small-object feature degradation as a coupled problem of representation, fusion, and prediction and indicates a practical accuracy-efficiency trade-off for dense aerial small-object perception.