image-features

A library for processing image features in a dataframe.

Dependencies :globe_with_meridians:

Python 3.11.6:

• pandas
• pyarrow
• imgembeddings
• pillow
• requests
• numpy
• torch

Raison D'être :thought_balloon:

image-features transforms URLs in a dataframe into image features.

Architecture :triangular_ruler:

image-features is a functional library, meaning that each phase of feature extraction gets put through a different function until the final output. The features its computes are as follows:

1. Embeddings - An embedding for each image in the row.
2. Luminosity - A measure of the luminosity of the image.
3. Mean Channels - The mean of each R/G/B channel.

Installation :inbox_tray:

This is a python package hosted on pypi, so to install simply run the following command:

pip install image-features

or install using this local repository:

python setup.py install --old-and-unmanageable

Usage example :eyes:

The use of image-features is entirely through code due to it being a library. It attempts to hide most of its complexity from the user, so it only has a few functions of relevance in its outward API.

Generating Features

To generate features:

import datetime

import pandas as pd

from imagefeatures.process import process

df = ... # Your timeseries dataframe
df = process(df, {"image_url"})

This will produce a dataframe that contains the new image related features.

License :memo:

The project is available under the MIT License.