image-classifier

Image Classifier NodeJS Package

Machine Learning Image Classifier for NodeJS

Installation

Library : npm install image-classifier

CLI : npm install -g image-classifier

ImageClassifier

Get Started

Create a new ImageClassifier with a dataset path for saving and loading your dataset. You can also specify to autosave to save every time you add a new example

    const { ImageClassifier } = require('image-classifier');

    const MyClassifier = new ImageClassifier('./mynewdataset.json', true);

Parameter	Description	Type	Mandatory
dataset	Path to dataset	String	True
autosave	Autosave dataset after adding a example (default: True)	Boolean	False

Load (Async)

Load should always be called before any further implementation. Will attempt to load dataset if the dataset path already exists

    MyClassifier.load().then(async function() {
        /* Add Examples here */
    });

Add Example (Async)

Add example data for training a new model

    // Add Toyota Exampless
    await MyClassifier.addExample('Toyota', './toyota0.png');
    await MyClassifier.addExample('Toyota', './toyota0.png');
    await MyClassifier.addExample('Toyota', './toyota0.png');
    /* ...Add more examples */

    // Add Honda Examples
    await MyClassifier.addExample('Honda', './honda0.png');
    await MyClassifier.addExample('Honda', './honda1.png');
    await MyClassifier.addExample('Honda', './honda2.png');
    /* ...Add more examples */

Parameter	Description	Type	Mandatory
label	Category label for what the image is	String	True
image	Path to image to add as an example for label	String	True

Predict (Async)

Get a classification prediction of image passed in

    const prediction = await MyClassifier.predict('./toyotaTest0.png');
    /* Prediction 
        {
            "classIndex": <Index of Label>,
            "label": <Label>,
            "confidences": {
                "Toyota": <Percentile>,
                "Honda": <Percentile>
            }
        }
    */

Parameter	Description	Type	Mandatory
image	Path to image for evaluating a prediction with your ImageClassifier	String	True

Drop Classifier (Sync)

Remove all examples of specified label

    MyClassifier.dropClassifier('Honda');

Parameter	Description	Type	Mandatory
label	Label for classifier you would like to remove from ImageClassifier. Drops all examples of specified label.	String	True

Save (Async)

Save dataset manually to dataset path passed from the constructor

    await MyClassifier.save();

train

WARNING: STILL IN BETA v1.0.0

Save

Save a dataset

train -s <someurl|somefile> <title.json>

Export

Export a dataset

train -e <somedataset.json> <newlocation.json>

Delete

Delete a dataset in your saved datasets

train -d <somedataset.json>

List

List all saved datasets

train -l

Identify

Identify an image with a saved dataset

train -i <someimage.png> <somedataset.json>

Help

Help menu will be brought up if none of the above flags are included

    Usage: train [options] <values>
            
    Options:
        -s, --save     : Save a dataset
        -e, --export   : Export a dataset
        -d, --delete   : Delete a dataset
        -l, --list     : List all saved datasets
        -i, --identify : Identify an image with a dataset
        -h, --help     : Open help menu
    
    Examples:
        train -s https://dataset.example.com/dataset.json example-dataset.json
        train -e example-dataset.json ./testing/new-example-dataset.json
        train -i cat.png example-dataset.json
        train -d example-dataset.json

Contributions