API Documentation Version 1.2

API Reference

Quick search

The Training API

The Training API consists of a set of urls used to interact with the IQ Engines computer vision database. At its core, this interface allows you to upload and train the IQ Engines computer vision system with objects using your own custom labels and metadata.

One way to try out the capabilities of the Training API without writing any code is to use the Training Console . We’ve developed this tool for you to quickly add objects into our computer vision database through a simple web interface.

For more advanced usage through a programmatic interface, please refer to the documentation below.

About REST (REpresentational State Transfer)

In order to expose the simplest interface possible to our computer visions database, we’ve designed the Training API using REST as a central guiding principle.

This means that the Training API is built up from a small set of primitive actions – CREATE, RETRIEVE, UPDATE, and DELETE – that operate over a defined set of resources which are addressable via URLs.

Core Data Model

The data in our computer vision system is represented using two core data types: objects and images. Objects describe real world entities using metadata and labels. Images point to the location of the raw pixel data. Many images may be associated to a single object

Here is an example of how we represent the object “coke can” and all of its associated images.

/static/apidocs/IQE_Data_Model.png

API Resources & Methods

/object/

Method Description More info
POST create a new object and related images create object

/object/<id>/

Method Description More info
GET retrieve object and its properties retrieve object details
DELETE delete object and related images delete object

/object/<id>/images/

Method Description More info
POST append images to the related object append to image list

/image/<id>/

Method Description More info
GET retrieves a given image and its attributes retrieve image details
DELETE delete particular image delete image