How does an image identified by Wolfram work?

The Image Identify by Wolfram project uses the highly automated "superfunctions" and as part of Wolfram Language API integration. It relies on a complex collection of meta-algorithms and built-in 'knowledge'. It has a built-in classifier trained from a large dataset using Wolfram Data Framework (WDF). However the main classifier is based on the deep neural networks. Source: How the Wolfram Language Image Identification Project Works The algorithm isn't perfect and misidentification is more likely to be caused by 'irrelevant objects repeatedly being in training images for a particular type of object'. You can read more at Wolfram Language Artificial Intelligence: The Image Identification Project.