Simuletic/CCTV_Incident_Dataset_Fall_Lying_Down_Detection

Overview

This is an open-source synthetic dataset for Computer Vision (CV) tasks, specifically designed for Fall Detection, Pose Estimation, and Incident Monitoring from overhead CCTV perspectives.

Unlike standard object detection datasets, this dataset includes Keypoints (Pose) annotations. This enables models to understand human posture and accurately distinguish between standing and fallen individuals.

🚀 Need more data? This is a sample dataset by Simuletic. We provide hyper-realistic synthetic data to solve edge cases in AI training. For larger, more diverse datasets or video sequences, visit simuletic.com/datasets.

Key Features

• Dual Annotations: Includes both Bounding Boxes and 17-Keypoint Skeletons (COCO standard layout).
• Targeted Scenarios: Focused heavily on incidents (approx. 95% of subjects are "laying"/fallen), making it ideal for training anomaly detection models.
• Privacy-First: Fully synthetic. No real individuals are depicted, eliminating GDPR/privacy risks.
• Compatibility: Ready for YOLOv8-Pose and YOLO11-Pose models.

Dataset Structure & Classes

The dataset follows the YOLO Pose format.

• images/: High-quality synthetic .jpg/.png files.
• labels/: .txt files containing class ID, bounding box, and keypoints.

Class Map

• 0: laying (Person lying down / fallen)
• 1: standing (Person standing / walking)

YAML Configuration

To use this with YOLO, your data.yaml should look like this:

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

Keypoints shape (17 points, x/y/visibility)

kpt_shape: [17, 3]

names: 0: laying 1: standing

Use Cases

Fall Detection: Train models to detect slip-and-fall accidents in real-time using pose analysis.

Healthcare Monitoring: Detect patients falling in hospitals or care homes (privacy-safe).

Public Safety: Identify medical emergencies or potential crime victims in surveillance feeds.

Ethics & License

Synthetic Nature: This data is computer-generated. No real humans were recorded or harmed.

License: CC BY 4.0. You are free to use and adapt this data, provided you give appropriate credit to Simuletic.

Citation

If you use this dataset in your research, please cite:

@dataset{simuletic_fall_detection_2025, author = {Simuletic Team}, title = {Simuletic Synthetic Fall & Incident Detection Dataset}, year = {2025}, url = {https://simuletic.com} }

Feedback? Reach out via https://simuletic.com or the "Issues" tab here on Kaggle.