ReadyML - Easy and Ready Machine Learning.

What is ReadyML?

ReadyML makes trained Machine Learning models ready to consume with minimum effort. With ReadyML you can to play with Image classification, Object Detection, Image Generation, Face Generation, Pose Detection... and much more!

This library is currently a prototype. If you face any issue or if the documentation is unclear, feel free to open a ticket issue

Model Categories

Name Description
Image Classification Provides category labels for an image
Object Detection Detect objects in a image
Image Generation Generate image from a category from scratch
Face Generation Generate a face image from scratch
Pose Detection Detect the keypoints of human pose (ankles, shoulders, elbows, ...)
Image Restoration
Image Super Resolution
Face Detection
Text Translation English to French, English to German, English to Russian, ...

Prerequisites

  • Python >= 3.7
  • The python package manager: pip

Installation

ReadyML installation is very convenient as it is packaged as a pip package. To install it, just run:

pip install readyml

pip will automatically install the dependencies like Tensorflow, Pytorch, Caffe...

How to use it?

The general format to use a model (infer) is that way:

from readyml import <model-category> as ric

# Initialize the model
model = ric.<TheModelIwantToUse>()

# Run the model
results = model.infer(<arguments>)

# Do something with the results

The format of the result differs from model to model. See the model names below to get a detailed description.

For example, to use the image classification model NASNetLarge, do:

from readyml import imageclassification as ric
import PIL.Image as Image

# Read an image
image_pil = Image.open("./images/greek_street.jpeg")

# Instantiate the model class
nasnetlarge = ric.NASNetLarge()
# Get and print the results
results = nasnetlarge.infer(image_pil)
print(results)

Results: The labels and their confidence score in percent.

[
    {
        "label": "monastery",
        "score": 38.63
    },
    {
        "label": "palace",
        "score": 21.18
    },
    {
        "label": "patio",
        "score": 14.9
    }
]

The above example means that the image can be categorized into three categories: - Monastery with a confidence score of 38.63%, - Palace with a confidence score of 21.18%, or - Patio with a confidence score of 14.9%,

Once ReadyML is intalled, you can start to use all the available models. Check the Models section