By Lionel Gibbons | January 24, 2017 | deep learning
According to a recent column from James Kobielus, published at InfoWorld, deep learning has taken machine learning from the traditional application to textual data and extended its reach to complex content types including video, voice, and music. It's what gives your smartphone's voice activated assistant the ability to understand you, and it's the technology behind voice and facial recognition. It provides the ability of a robot or a drone to respond intelligently to the environmental conditions that surround it. But, Kobielus notes, deep learning goes even farther. It alters reality.
Deep learning is advanced enough at this point that a digital image can be perfected by superimposing visual elements that were missing in the original photo. It grants the ability for a computer to generate an authentically sounding human voice. To create artwork so that it appears to have been composed by a human artist in a specific style. Deep learning can transform a sketch into a photorealistic image.
The benefits of deep learning are many, Kobielus' column asserts, but so are the risks. If these algorithms are used to operate self-driven vehicles or to build prosthetic limbs, there must be transparency in how the algorithm's path was determined. And there should, states the article, also be some way to let the public know when the reality it is experiencing has been modified by deep learning.
Are you looking to build an enterprise grade deep learning environment? Bright can provide everything you need to get your deep learning environment up and running, and manage it. We'll give you a choice of frameworks and Machine Learning libraries, as well as the modules to support it. Get in touch to find out more.