After learning poetry, Google AI now intend to challenge music, painting and the original video

Updated: May 23, 2016  Views: 195

After learning poetry, Google AI now intend to challenge music, painting and the original video


SAN FRANCISCO, May 23 pm news on Moogfest music and technology festival, held last week, Google's artificial intelligence department Google Brain researcher Ike Douglas (Douglas Eck) announced that it will launch a new project, hoping to find out whether we can have a real computer creativity.

The project called Magenta will be publicly released in early June this year, but has been able to advance Moogfest participants learn some information .Magenta using Google developed TensorFlow machine learning engine (the engine was open by the end of 2015), to determine artificial intelligent system can be trained to create original music, painting or video.

This task is not simple, because even the most advanced artificial intelligence system, it is difficult to imitate the style of painters and artists, let alone that they themselves created. Ike recognition, artificial intelligence systems have long narrative is far from capacity.

But Magenta hope to develop some tools to help other researchers and their teams to explore the creative potential of the computer, just like before TensorFlow open source code, like Ike represents, Magenta will also be open to the public their own tools.

They first thing to do is to develop a program to help researchers to import music data TensorFlow use MIDI music files to the system to expand the training of music knowledge.

Ike as a member of the team, Adam Roberts (Adam Roberts) represent, Magenta project will be June 1 release more resource-related information, and add new software on its GitHub page, but also regularly blog posted on the latest developments.

Roberts also showed he is developing a simple digital synthesizer program that allows artificial intelligence can understand his notes that are playing and re-playing a more complete melody.

Ike said the goal of the system is to develop a set of melodies to regularly create new systems in order to bring music to relax on, so that it can sit on the couch at home listening to the wonderful music creation computer. But he also admitted, at least for the foreseeable future, this art still requires a certain degree of human intervention, to make the robot independently creative process is indeed very difficult.

Ike representation, Magenta inspired by other projects Google Brain, such as Google DeepDream, the purpose of the project is to train artificial intelligence system image database 'to fill the vacancy in the picture', trying to find some this may not be rendered by drawing itself out painting structure. the system will eventually create some illusion picture, in the conventional picture added skyscrapers, eye and daily necessities.

In Magenta system, Ike and his team will want to know, whether through adequate training for machine data to create compelling music. Ike said that after solving the problem of music, they are also trying to create images and video.

Although Google artificial intelligence system in the short term is unlikely to replace top artists, but Ike believes that there may be applied to certain areas of computer-generated music in the short term. For example, if someone with a wearable device to track your heartbeat, to intelligent phone indicates that the person sending nervous signals, the creation of artificial intelligence systems to soft music can help relieve stress. (book Yu) Get Top News, Latest News and Current News from Latest Current News: U.S., China, World, Entertainment, Health, Finance, Business, echnology, Travel, Politics, Sports.

Traditional Chinese: 學會作詩後,谷歌AI現在打算挑戰音樂、繪畫和視頻原創了

Related Technology