This is a very simple program to demonstrate the Segment Anything capability.
According to the Segment Anything documentation, the code requires python>=3.8, as well as pytorch>=1.7 and torchvision>=0.8. This code has only been tested with 'python3.11', but should work on 'python3.8'. NOTE: python3.11 is the latest version to support Torch.
To run this within a Python virtual environment (venv), create the venv:
$ python -m venv tobiiSAMenv
Activate the venv
$ source tobiiSAMenv/bin/activate
Install the dependencies:
$ pip install numpy
$ pip install torch
$ pip install matplotlib
$ pip install opencv-python
$ pip install segment-anything-py
Download the model:
$ mkdir model
$ cd model
$ curl -O https://dl.fbaipublicfiles.com/segment_anything/sam_vit_h_4b8939.pth
$ cd ..
Run the program
$ python predictor_example.py