CCTV Weapon Detection Dataset




This synthetic weapon detection dataset is built to train AI powered weapon detection software the way it actually gets deployed — on real CCTV feeds. Every image is computer-generated from realistic security-camera angles (overhead, corner-mount, doorway, corridor) and includes people carrying and drawing weapons in natural indoor and outdoor environments: retail stores, parking lots, hallways, transport hubs, and public spaces. Because the dataset is 100% synthetic, it covers scenarios that are almost impossible to capture responsibly with real footage — including AI concealed weapons detection cases where handguns are partially hidden behind jackets, bags, or the person's own body. Lighting spans bright daylight to dim nighttime, with realistic motion blur, sensor noise, compression artifacts, and occlusion. Each image is annotated in YOLO format with bounding boxes for both 'person' and 'weapon' classes, ready to plug into YOLOv5, YOLOv8, or YOLOv11 training pipelines. The open-source sample includes 150 images; the full commercial package contains 1,000+ images with 15 evaluation videos for benchmarking real-time detection performance.
1,000+ images
Full Package
150
Open Source Samples
YOLO
Annotation Format
100%
Privacy Compliant
Dataset Features
Intended Use Cases
Free Sample vs. Commercial Package
Free Open-Source Sample
- 150 annotated images
- Format: YOLO
- Hosted on Kaggle
- Hosted on Hugging Face
- Licence: See the hosting platform's terms of use
Commercial Package
- 1,000+ images
- Format: YOLO
- Licence: Student & Research or Business licence
The Creative Commons licence above applies only to the free sample, not to the full commercial package.
Limitations & Recommended Validation
This dataset is 100% synthetic. While it is designed to closely match real-world sensor and camera conditions, synthetic imagery can still differ from live footage in ways that affect model accuracy (a "domain gap"). Validate a trained model against real-world footage from your specific deployment environment before production use.
Not intended as a sole basis for biometric identification, legal evidence, or safety-critical decisions without independent human review and real-world testing.