The Fast Segment Anything Model (FastSAM) is a real-time CNN-based model that can segment any object within an image based on various user prompts. Segment Anything task is designed to make vision tasks easier by providing an efficient way to identify objects in an image. FastSAM significantly reduces computational demands while maintaining competitive performance, making it a practical choice for a variety of vision tasks.
The tutorial consists of the following steps:
- Install and import prerequisite packages
- Download the Fast Segment Anything Model using the Ultralytics package.
- Run the unconditioned segmentation mask generation pipeline
- Convert the model backing the FastSAM pipeline
- Quantize the model using NNCF
- Run interactive segmentation pipeline using OpenVINO and Gradio
This is a self-contained example that relies solely on its own code.
We recommend running the notebook in a virtual environment. You only need a Jupyter server to start.
For details, please refer to Installation Guide.