EDNet: Edge-Optimized Small Target Detection in UAV Imagery -- Faster Context Attention, Better Feature Fusion, and Hardware Acceleration

TL;DR

EDNet is a lightweight, edge-optimized detection framework based on YOLOv10, designed for real-time small target detection in UAV imagery, achieving high accuracy and speed across various devices.

Contribution

The paper introduces EDNet, a novel edge-optimized detection architecture with enhanced feature fusion and context attention, tailored for real-time small target detection in drone imagery.

Findings

Up to 5.6% mAP@50 improvement over baseline

Operates at 16-55 FPS on an iPhone 12

Supports seven model sizes for diverse deployment

Abstract

Detecting small targets in drone imagery is challenging due to low resolution, complex backgrounds, and dynamic scenes. We propose EDNet, a novel edge-target detection framework built on an enhanced YOLOv10 architecture, optimized for real-time applications without post-processing. EDNet incorporates an XSmall detection head and a Cross Concat strategy to improve feature fusion and multi-scale context awareness for detecting tiny targets in diverse environments. Our unique C2f-FCA block employs Faster Context Attention to enhance feature extraction while reducing computational complexity. The WIoU loss function is employed for improved bounding box regression. With seven model sizes ranging from Tiny to XL, EDNet accommodates various deployment environments, enabling local real-time inference and ensuring data privacy. Notably, EDNet achieves up to a 5.6% gain in mAP@50 with significantly fewer parameters. On an iPhone 12, EDNet variants operate at speeds ranging from 16 to 55 FPS, providing a scalable and efficient solution for edge-based object detection in challenging drone imagery. The source code and pre-trained models are available at: https://github.com/zsniko/EDNet.