LAF-YOLOv10 with Partial Convolution Backbone, Attention-Guided Feature Pyramid, Auxiliary P2 Head, and Wise-IoU Loss for Small Object Detection in Drone Aerial Imagery

TL;DR

This study introduces LAF-YOLOv10, built on YOLOv10n, integrating four complementary techniques to improve small-object detection in drone imagery, integrating four complementary techniques to improve small-object detection in drone imagery.

Abstract

Unmanned aerial vehicles serve as primary sensing platforms for surveillance, traffic monitoring, and disaster response, making aerial object detection a central problem in applied computer vision. Current detectors struggle with UAV-specific challenges: targets spanning only a few pixels, cluttered backgrounds, heavy occlusion, and strict onboard computational budgets. This study introduces LAF-YOLOv10, built on YOLOv10n, integrating four complementary techniques to improve small-object detection in drone imagery. A Partial Convolution C2f (PC-C2f) module restricts spatial convolution to one quarter of backbone channels, reducing redundant computation while preserving discriminative capacity. An Attention-Guided Feature Pyramid Network (AG-FPN) inserts Squeeze-and-Excitation channel gates before multi-scale fusion and replaces nearest-neighbor upsampling with DySample for content-aware interpolation. An auxiliary P2 detection head at 160$\times$160 resolution extends localization to objects below 8$\times$8 pixels, while the P5 head is removed to redistribute parameters. Wise-IoU v3 replaces CIoU for bounding box regression, attenuating gradients from noisy annotations in crowded aerial scenes. The four modules address non-overlapping bottlenecks: PC-C2f compresses backbone computation, AG-FPN refines cross-scale fusion, the P2 head recovers spatial resolution, and Wise-IoU stabilizes regression under label noise. No individual component is novel; the contribution is the joint integration within a single YOLOv10 framework. Across three training runs (seeds 42, 123, 256), LAF-YOLOv10 achieves 35.1$\pm$0.3\% mAP@0.5 on VisDrone-DET2019 with 2.3\,M parameters, exceeding YOLOv10n by 3.3 points. Cross-dataset evaluation on UAVDT yields 35.8$\pm$0.4\% mAP@0.5. Benchmarks on NVIDIA Jetson Orin Nano confirm 24.3 FPS at FP16, demonstrating viability for embedded UAV deployment.

Figures (5)

• Figure 1: LAF-YOLOv10 architecture. PC-C2f backbone (blue), SE-gated AG-FPN neck (purple), and P2/P3/P4 detection heads (green); P5 removed, Wise-IoU v3 ref_wiou used for regression.
• Figure 2: Internal structure of (a) PC-C2f block ref_pconv with $C/4$ active channels and (b) AG-FPN fusion via SE gating ref_se and DySample ref_dysample upsampling.
• Figure 3: Speed--accuracy trade-off on VisDrone-DET2019 ref_visdrone (RTX 4090, FP16, batch 1).
• Figure 4: Training convergence on VisDrone-DET2019 ref_visdrone (seed 42, mAP@0.5 every 5 epochs).
• Figure 5: Per-category mAP@0.5 on VisDrone-DET2019 ref_visdrone. Geometrically complex classes benefit most.