Datasets:
image imagewidth (px) 784 784 |
|---|
Weapon Detection Dataset: Rifles vs Umbrellas (CCTV, YOLO): free 120-image sample
This weapon detection dataset is a free sample of 120 synthetic CCTV images of people carrying rifles or closed umbrellas, with YOLO bounding boxes for person, rifle and umbrella, for training and testing gun detection models that do not mistake umbrellas for rifles. The full 1,000-image dataset keeps the balanced 50 / 50 rifle-versus-umbrella split and adds daylight, dusk and night scenes, weather, sensor noise, motion blur and 8 evaluation videos. Made by Simuletic, a Swedish company that has built synthetic CCTV datasets for computer vision since 2023.
Get the full dataset · Free sample licence: CC BY-NC 4.0. This free sample is also on Kaggle.
Facts
| Free sample (this repo) | Full dataset | |
|---|---|---|
| Images | 120 JPEG | 1,000, plus 8 evaluation videos |
| Classes | person, rifle, umbrella |
person, rifle, umbrella |
| Boxes | 138 person, 64 rifle, 59 umbrella | every person, rifle and umbrella boxed |
| Hard negatives | umbrellas as a labelled class (about half the images) | 50%: umbrellas as a labelled class |
| Resolution | 784 × 1,168 px | CCTV frames; daylight, dusk and night |
| Camera view | high-angle and wide-angle CCTV; streets, stairwells, public spaces | varied weather, sensor noise and motion blur |
| Labels | YOLO .txt |
YOLO, dataset card |
| Licence | CC BY-NC 4.0 | one-time: research $149 / commercial $590 per 1,000 images |
What's in the images
People carrying either a rifle or a closed umbrella, seen from high-angle and wide-angle CCTV cameras in streets, stairwells and other public spaces. Seen from a CCTV camera, a rifle and a closed umbrella have a similar long, thin shape, which makes umbrellas a common cause of false alarms in weapon detection. Training on both (hard negative mining) teaches a model the differences in shape, grip and posture. 64 images contain a rifle and 58 an umbrella, a near 50/50 split. Rifles are assault rifles and carbines held in different ways; umbrellas are closed black umbrellas held by the stem or handle. The images include sensor noise, motion blur and varied lighting.
Classes and boxes
| Class | YOLO id | Boxes in sample |
|---|---|---|
person |
0 | 138 |
rifle |
1 | 64 |
umbrella |
2 | 59 |
Label format and folder layout
Simuletic_Weapon_Umbrella_Dataset/
data.yaml # class config (see note below)
Example_image_weapon.JPG # preview images (not labelled)
Example_image_umbrella.JPG
images/ # 120 JPEG images (person_rifle_vs_umbrella_001.JPG, ...)
labels/ # one YOLO .txt per image
Each label line is class_id x_center y_center width height, normalised to 0–1:
0 0.531 0.668 0.482 0.588
2 0.747 0.776 0.084 0.267
Quick start (Ultralytics YOLO)
Create a data.yaml next to the Simuletic_Weapon_Umbrella_Dataset/ folder:
path: Simuletic_Weapon_Umbrella_Dataset
train: images
val: images # small sample: use the same images, or split them yourself
names:
0: person
1: rifle
2: umbrella
The included data.yaml names class 1 weapon and says nc: 1, which Ultralytics rejects because three names are listed; use the file above instead.
Then train:
from ultralytics import YOLO
model = YOLO("yolo11n.pt")
model.train(data="data.yaml", epochs=30, imgsz=640)
Use cases
- Reducing false alarms from umbrellas in weapon detection
- Rifle detection for schools, airports, perimeters and public spaces
- Testing whether an existing detector confuses umbrellas with rifles
- Pre-training before fine-tuning on your own camera footage
Full dataset and prices
The full CCTV Weapon Detection: Rifles vs Umbrellas has 1,000 images, YOLO, classes person, rifle, umbrella. Every licence is a one-time payment, perpetual, for one organisation; you get a download link within five working days (card or invoice). Research covers students and non-commercial work and is available for 1,000 images only; packs of 5,000 images and more are sold with the commercial licence, which lets you train models and ship them in products.
| Research | Commercial | |
|---|---|---|
| 1,000 images | $149 | $590 |
| 5,000-image pack | not offered | $2,400 |
| 10,000-image pack | not offered | $4,500 |
| 20,000-image pack | not offered | $8,500 |
Packs are the same dataset at larger size, with the same labels and split, under the commercial licence. 50,000+ images, video datasets and enterprise agreements (from $10,000) are quoted. Other scenes, camera angles or classes: custom data or contact@simuletic.com.
How the images were made
The images are fully synthetic (computer-generated) by Simuletic: no real people, places or incidents are depicted, so there are no privacy or consent issues.
Limitations. Synthetic images do not cover every camera, lens, compression artefact or scene you will meet in deployment. Validate on your own footage, and for production combine synthetic and real data.
More synthetic datasets from Simuletic
- CCTV Weapon Detection Dataset
- Weapon Detection: Handgun vs Bag of Chips
- CCTV Knife Detection Dataset
- CCTV ATM Robbery Detection Dataset: Gun & Knife
- All datasets: simuletic.com/datasets · Hugging Face · Kaggle · GitHub
Licence
This free sample is licensed under CC BY-NC 4.0: you may share and adapt it for non-commercial purposes, as long as you credit Simuletic. The full dataset is sold under a separate one-time research or commercial licence; it is not redistributed.
Citation
@misc{simuletic_rifles_umbrellas_2026,
author = {{Simuletic}},
title = {CCTV Weapon Detection: Rifles vs Umbrellas},
year = {2026},
publisher = {Simuletic},
howpublished = {\url{https://huggingface.co/datasets/Simuletic/CCTV_Weapon_Detection_Rifles_vs_Umbrellas}},
note = {Free sample. Full dataset: \url{https://simuletic.com/datasets/cctv-weapon-detection-rifles-vs-umbrellas}}
}
Questions or feedback: contact@simuletic.com
- Downloads last month
- 952