Demo Streamlit Few Shot Learning

Showcasing the power of streamlit to display a demo of a Few Shot Learning model using keras-fsl. The goal of this project is to help you try streamlit.

Install

You will need :

python 3.8
preferably a venv
poetry 1+ : curl -sSL https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py | python -
install dependencies in the venv : poetry install in the root folder

What to look at

Try yourself

Go to simple_few_shot_classifier.py and try to transform it in a streamlit app. Run your app : PYTHONPATH=. streamlit run simple_few_shot_classifier.py

Some Demos

A few shot learning streamlit app :

using the images in the data/default_catalog folder as catalog and data/images as images to predict example/streamlit_simple_few_shot_classifier.py
using the images in the data/default_catalog folder as catalog and predicting images from user input example/streamlit_simple_few_shot_classifier_with_input_from_user.py
getting images to predict and to add to the catalog from the user example/streamlit_simple_few_shot_classifier_with_catalog_from_user.py

Wirg/demo-streamlit-fsl

Demo Streamlit Few Shot Learning

Install

What to look at

Try yourself

Some Demos