/yolo_world_video

This interface allows users to upload videos, select object detection parameters, and visualize the processed video with detected objects highlighted.

Primary LanguagePythonApache License 2.0Apache-2.0

title emoji colorFrom colorTo sdk sdk_version app_file pinned license
Yolo World Video
🌎
yellow
pink
gradio
4.19.2
app.py
false
apache-2.0

🌎 YOLO-World Video Processing on Hugging Face Spaces

Welcome to our YOLO-World video processing project on Hugging Face Spaces! This project leverages the power of YOLO (You Only Look Once) models for efficient and accurate real-time object detection in videos. Our interface allows users to upload videos, select object detection parameters, and visualize the processed video with detected objects highlighted.

Features

  • Video Upload: Users can upload their videos for object detection.
  • Model Selection: Choose between different YOLO models (yolov8s-world.pt, yolov8m-world.pt, yolov8l-world.pt) for varying levels of accuracy and processing speed.
  • Custom Object Detection: Enter specific categories for detection to tailor the model to your needs.
  • Adjustable Confidence and IoU Thresholds: Fine-tune the detection sensitivity and intersection-over-union thresholds for optimal accuracy.
  • Real-Time Progress: Track the processing progress with a real-time progress bar.

How It Works

  1. Upload a Video: Begin by uploading a video file that you want to process.
  2. Set Parameters: Enter the categories you're interested in detecting (comma-separated), select a YOLO model, and adjust the confidence and IoU thresholds as needed.
  3. Process Video: Click the "Process video" button to start the object detection process. The system will analyze each frame of the video, detect objects based on your parameters, and highlight them.
  4. View Results: Once processing is complete, the output video will be displayed, showing the detected objects with bounding boxes.

Technologies Used

  • Gradio: For creating the interactive web interface.
  • YOLO (You Only Look Once): For real-time object detection.
  • OpenCV: For video processing and rendering.
  • Hugging Face Spaces: Hosting the interactive application.

Local Setup (Optional)

If you prefer to run this project locally, follow these steps:

  1. Clone the repository: git clone https://github.com/your-repo/yolo-world-space.git
  2. Install dependencies: pip install -r requirements.txt
  3. Run the application: python app.py
  4. Open your web browser and navigate to the URL displayed in your terminal.

Contribute

We welcome contributions! Whether it's improving the detection algorithm, enhancing the interface, or fixing bugs, feel free to fork the repository and submit a pull request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

  • The YOLO authors for their groundbreaking work in real-time object detection.
  • The Gradio and Hugging Face teams for their amazing tools that make deploying AI apps easier.

Happy Detecting!