Pupil Labs AOIs Module: Automate AOI Definition with GroundingSAM

Introduction

The Pupil Labs AOIs Module leverages the powerful GroundingSAM framework to automate the process of defining Areas of Interest (AOIs) within the Reference Image Mapper/Marker Mapper enrichment in Cloud. This innovative tool offers a user-friendly web interface, allowing for the upload of reference images and the specification of desired segmentation parameters. By utilizing GroundingSAM for segmentation, and facilitating mask submission through our Cloud API, this module significantly streamlines the enrichment process, eliminating the need for manual mask drawing.

Quick Start with Google Colab

Local Installation

Docker Installation

1. Clone the repository and build the Docker image:

git clone https://github.com/pupil-labs/aois_module
cd aois_module
docker build -t pupil-labs-aois-module .

Python Package Installation

Optional step to set up a virtual environment using UV (from astral.sh):

curl -LsSf https://astral.sh/uv/install.sh | sh
uv venv --seed --python3.11
source .venv/bin/activate
uv pip install pupil-labs-aois-module

Alternatively, install directly with pip:

pip install pupil-labs-aois-module

Or, for a development setup, clone the repository and install:

git clone https://github.com/pupil-labs/aois-module.git
cd aois-module
pip install .

Running the Module

Locally

pl-aois

Using Docker

docker run -p 8002:8002 pupil-labs-aois-module