Important: This documentation covers Yarn 1 (Classic).
For Yarn 2+ docs and migration guide, see yarnpkg.com.

Package detail

imagine-client-js

doc-ai3ISC0.2.4

A Nodejs client for doc.ai's Imagine API

readme

Imagine Client for NodeJS

This library allows you to easily connect to doc.ai's Imagine API.

Supported Image Models

Model Id Type Description
phenomenal-face image Given a facial image, this model infers the age, sex, height and weight of a person
happy-face image Given a facial image, this model infers the mood of a person

API


> Given a string (modelId), gets the image model. See the *"Supported Image Models"* section for a list of supported model ids

```imageModel.infer(fsReadStream, shouldStoreImage, metadata)

Performs the model's inference on an image. A read stream should be passed as the parameter (fsReadStream).

imageModel.inferByType(data, type, shouldStoreImage, metadata)

Performs the model's inference on an image. A data should be passed as the parameter (data) and type should specify the format that the data is in. (base64 is currently supported)

imageModel.correctionLabel(correctionObj, metadata)

Uploads a correction label to the imagine API, for a given inference.

imageModel.predictionLabel(correctionObj, metadata)

(FOR EDGE PREDICTIONS) Uploads a prediction label to the imagine API, for a given inference.

Example


const fs = require('fs');
const ImagineClient = require('imagine-client-nodejs');

// Initialize imagine client
const client = new ImagineClient({
  apiKey: '1234567890',
  clientId: '3ad839e3-90e7-4564-b3a1-5e39c88545a5'
});

// Get a reference to the 'happy-face' model
const model = client.getImageModel('happy-face');

// Create a read stream to a image file
const imageStream = fs.createReadStream('face.png');

// Perform the model's inference, on the image.
return model.infer(imageStream, false)
  .then((result) => {
    // Print the result
    console.log(result);
  });