Multiple Overlapping Object Detection

This project demonstrates the usage of a pre-trained model in TensorFlow.js for performing multiple (overlapping) object detection in images. It utilizes the COCO-SSD model to recognize objects and their positions with high accuracy.

Note: The npm files included in this project are not necessary for the code to work currently. They are intended for the future migration of this project to MediaPipe.

Setup

1. Clone the repository:

git clone https://github.com/octopols/Multiple-Overlapping-Object-Detection.git

2. Navigate to the project directory:

cd Multiple-Overlapping-Object-Detection

3. Open the project in a web browser:
   • You can use a local development server such as Live Server or serve the files using Python's built-in HTTP server. For example, using Python 3:

python3 -m http.server

4. Access the application:
   • Open your web browser and visit http://localhost:8000 (or the appropriate port number if using a different local server).

Credits

This project utilizes TensorFlow.js and the COCO-SSD model. The code was initially based on a demo by Jason Mayes and was later modified by Hirnaymay Bhaskar.

• TensorFlow.js
• COCO-SSD Model

Contributing

Contributions are welcome! If you have any ideas, suggestions, or bug reports, please open an issue or submit a pull request.

License

This project is licensed under the MIT License.