isaoselmisdev/Falldetection

FallDetection - Real-Time Fall Detection

This project implements a real-time fall detection system using a YOLO-based object detection model. The system is designed to work on Hugging Face Spaces with Docker and Gradio for an interactive user experience.

🚀 What It Does

• —Accepts a .mp4 video file as input.
• —Detects falls in real-time using the YOLO model (best.pt).
• —Identifies and labels different states such as:
• —Falling
• —Sitting
• —Confirmed Fallen
• —Outputs a processed video with bounding boxes and labels indicating the detected status.

🛠️ Technologies Used

• —Gradio: For creating an easy-to-use web interface.
• —YOLO (Ultralytics): For detecting and classifying fall-related activities.
• —OpenCV: For processing video frames and generating the output.
• —Docker: For deploying the app on Hugging Face Spaces.