Datasets:

Simuletic
/

UAV-Aerial-View-Battle-Tank-Detection-Dataset

This is an open-source synthetic dataset specifically engineered to train computer vision models in identifying main battle tanks (MBTs) and armored combat vehicles from tactical overhead and drone perspectives.

Obtaining real-world tactical aerial imagery for defense analytics is heavily restricted, operationally dangerous, or classified. This dataset addresses that critical data bottleneck by providing high-fidelity, high-angle aerial simulation frames across a diverse array of operational environments—including harsh deserts, complex Eastern European villages, and open fields.

🚀 Need a comprehensive enterprise deployment? This is a verified sample dataset curated by Simuletic. We engineer high-fidelity synthetic datasets to solve edge cases, severe environmental occlusions, and rare tactical vulnerabilities in AI training. For customized camera pitches, specific vehicle types, thermal infrared (FLIR) sensory layers, or expanded multi-sensor simulation, visit simuletic.com/datasets.

Key Features

UAV Perspective Realism: Simulates true-to-life tactical drone flight heights, camera focal lengths, atmospheric haze, and varying look-angles (oblique to 90-degree nadir top-down).
Multi-Theater Environment Randomization: Features armored assets deployed across visually distinct operational backdrops, including dry arid desert terrain, dense rural Eastern European villages, and seasonal vegetation fields.
Mathematical Annotation Precision: Programmatically verified spatial annotations that accurately bound the active target asset, accounting perfectly for shadow orientation, track orientation, and gun barrel pitch.
100% Secure & Frictionless: Generated entirely using simulation pipelines. Contains zero classified sensor signatures, restricted airspace imagery, or geolocational liabilities.
YOLO Optimized: Fully structured and pre-annotated in standard normalized YOLO bounding box (.txt) format, ready for immediate training pipelines.

Dataset Structure & Classes

The dataset repository matches standard object detection conventions:

images/: High-fidelity synthetic aerial drone image files.
labels/: Normalized .txt files containing target class indices and bounding box anchors.

Class Map

0: battle_tank (Main battle tanks and heavy tracked armored vehicles)

YAML Configuration

To initialize training with Ultralytics YOLO architectures, your data.yaml layout should be configured as follows:

path: /path/to/dataset
train: images
val: images

nc: 1
names:
  0: battle_tank

Core Use Cases

    Autonomous Reconnaissance & Tracking: Power edge AI models inside tactical unmanned aerial systems to trigger immediate threat updates when a tracked vehicle enters an asset boundary.

    Defense Threat Assessment: Train deep neural networks to recognize combat assets camouflaged or positioned near civilian architecture, natural tree lines, or deep desert dunes.

    Sim-to-Real (Sim2Real) Research: Provide clean, geometrically perfect baselines for advanced domain adaptation testing, enabling robust translation to edge-degraded optical feeds.

Ethics & License

    Synthetic Nature: This data is 100% computer-generated. No real military maneuvers were monitored, recorded, or compromised in the creation of this project.

    License: CC BY 4.0. You are free to share, adapt, and build upon this data for academic research or commercial production pipelines, provided appropriate attribution credit is given to Simuletic.

Citation

If you utilize this dataset in your models, safety evaluations, or publications, please cite the framework as follows:
Kodavsnitt

@dataset{simuletic_uav_tank_detection_2026,
  author = {Simuletic Team},
  title = {Simuletic Aerial UAV Perspective Battle Tank Detection Dataset},
  year = {2026},
  url = {[https://simuletic.com](https://simuletic.com)}
}

Feedback & Collaboration

Looking for custom operational scene layouts, distinct armored asset variants, or specific multi-spectral sensor configurations? Reach out directly via simuletic.com or open a ticket in the Discussion tab here on Kaggle!