The Google Brain's Team latest project aims to use TensorFlow's machine learning capabilities to address the question, “Can we use machine learning to create compelling art and music?". The first music clip released by the Magenta team, suggests there is a still some way to go.

It is the piano tune in this 90 second clip that is the AI-generated composition, primed with the four notes C,C,G,C. The drum beat was added by researchers "to give context to the computer's harmonic rhythm". The sample is described as:

an interesting amalgamation of sounds: simple, but full of complex musical ideas like repeated phrasing, form, and feeling.

This may be a bit of an overstatement but it is a start and we've seen other machine learning projects go from small starts to more impressive results. 

If you just listen to the start you might conclude that there is a long way to go but the second half of the piece is much more interesting. It doesn't quite go where a human might take it but this in itself is intriguing. Just as it finishes you feel that it might be just about to "take off". 

Douglas Eck, the project leader, posted Welcome to Magenta on its website on Jume 1st but had disclosed rather more about the project at Moogfest the previous week. He revealed that the inspiration for Magenta had come from other Google Brain projects, like Google DeepDream, where AI systems were trained on image databases to “fill in the gaps” in pictures, trying to find structures in images that weren’t necessarily present in the images themselves. In the Magenta project Eck and his team want to see if, given enough training data, a machine could create music that would be engaging and exciting for a person to listen to.  Eck also suggested that computer-created music could be used therapeutically, suggesting a scenario in which a person’s wearable device tracking her heart rate sends a signal to her smartphone that she is stressed and an AI system could generate appropriately soothing music to alleviate the stress.  

The Magenta GitHub repository states that:

Magenta encompasses two goals. It’s first a research project to advance the state-of-the art in music, video, image, and text generation. Second, it's an attempt to build a community of artists, coders, and machine learning researchers.

This repo only opened 5 days ago, and has already been forked 119 times which suggests that there a good deal of interest. 

If you want to get involved with Magenta it is 

Magenta project, an effort using machine intelligence. Magenta is part of the Google Brain team and is using TensorFlow (www.tensorflow.org), an open-source library for machine learning. The question Magenta asks is,

“Can machines make music and art? If so, how? If not, why not?”

The goal of Magenta is to produce open-source tools and models that help creative people be even more creative. I’m primarily looking at how to use so-called “generative” machine learning models to create engaging media. Additionally, I’m working on how to bring other aspects of the creative process into play. For example, art and music is not just about generating new pieces. It’s also about drawing one’s attention, being surprising, telling an interesting story, knowing what’s interesting in a scene, and so on.

More Information

Magenta website

Magenta on GitHub 

Related Articles

TensorFlow - Googles Open Source AI And Computation Engine

TPU Is Google's Seven Year Lead In AI

No Glittering Prizes For Creative Robots

Can Robots Create Art? 

To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook, Google+ or Linkedin.

Comments

Make a Comment or View Existing Comments Using Disqus

or email your comment to: comments@i-programmer.info