GewisLab/AMAD-5-dataset
数据集
数据集查看器
文件和版本
Pull Requests
讨论

AMAD-5: Aerial Military Asset Detection Dataset

5-class curated dataset for aerial military asset detection in UAV surveillance.

数据集概述 / Overview

AMAD-5 是一个精选的无人机视角航拍数据集,包含五类军事与民用目标。该数据集专为航拍监控中的目标检测任务设计,涵盖车队、混合地形及平民场景等真实多目标环境。每张图像均提供 YOLO 格式 的边界框标注,可直接用于目标检测任务,也可适配图像分类任务。

AMAD-5 is a curated aerial dataset containing images of five classes of military and contextual assets captured from a UAV perspective. The dataset is designed for object detection in aerial surveillance scenarios and reflects real-world multi-object environments including convoys, mixed terrain, and civilian presence. Each image is annotated with bounding boxes in YOLO format, making the dataset directly usable for object detection tasks and adaptable for image classification if required.

类别 / Classes

Class IDLabel说明
0military_tank军事坦克
1military_vehicle军用车辆(含多种轮式装备)
2civilian平民
3soldier士兵
4civilian_vehicle民用车辆

military_vehicle 类整合了多种轮式军事装备,确保航拍检测场景下的视觉一致性。

数据集统计 / Dataset Statistics

数据划分 / Data Split

Split图像数 / Images标签文件数 / Label Files大小 / Size
Train27,58328,2613,267.1 MB
Val4,1844,268451.5 MB
Test2,3452,431268.9 MB
Total34,11234,960~4,683 MB (4.7 GB)

标注统计 / Annotation Statistics

指标 / Metric数值 / Value
总标注框 / Total Bounding Boxes70,704
每张图像目标数(最小)1
每张图像目标数(最大)86
每张图像目标数(平均)2.0
每张图像目标数(中位数)1

类别分布 / Class Distribution

Class IDLabel标注数 / Count占比 / Percentage
0military_tank20,84529.5%
1military_vehicle15,87322.4%
2civilian13,37218.9%
3soldier11,77316.7%
4civilian_vehicle8,84112.5%

目录结构 / Directory Structure

├── train/
│   ├── images/      # 27,583 张训练图像
│   └── labels/      # 28,261 个 YOLO 标注文件
├── val/
│   ├── images/      # 4,184 张验证图像
│   └── labels/      # 4,268 个 YOLO 标注文件
├── test/
│   ├── images/      # 2,345 张测试图像
│   └── labels/      # 2,431 个 YOLO 标注文件
├── dataset.yaml     # 数据集配置文件
└── README.md

标注格式 / Annotation Format

所有标注文件遵循标准 YOLO 格式:

class_id x_center y_center width height
  • 坐标值已归一化到 [0, 1] 区间
  • 兼容 YOLOv5 / YOLOv8 / YOLOv11 / Faster R-CNN / EfficientDet 等框架

兼容框架 / Compatible Frameworks

  • 目标检测: YOLOv5, YOLOv8, YOLOv11, Faster R-CNN, EfficientDet
  • 图像分类: ResNet, EfficientNet, MobileNet

致谢 / Acknowledgements

  • 感谢 AtomGit 平台 提供数据存储与托管服务。
    Special thanks to AtomGit for providing data storage and hosting services.
  • 本数据集由 gewis lab 非官方组织上传维护。
    This dataset is uploaded and maintained by the unofficial organization gewis lab.
  • 数据集版权归原作者所有,如需查找原始作者及获取更多信息,请访问 Kaggle 数据集主页:
    AMAD-5: Aerial Military Asset Detection Dataset on Kaggle
  • 隶属于 天津商业大学 (TJUCE) AMAD 项目系列。
    Affiliated with the AMAD project series of Tianjin University of Commerce (TJUCE).

免责声明: 本数据集仅用于学术研究目的,请遵守相关法律法规及数据使用条款。 Disclaimer: This dataset is intended for academic research purposes only. Please comply with applicable laws, regulations, and data usage terms.