Google’s first AI-powered doodle is an adorable tribute to Bach

Google’s first AI-powered doodle is an adorable tribute to Bach that lets users create their own harmonized melodies

  • Google developed an AI model that was trained on 306 of Bach’s harmonizations
  • This allows users to compose their own two-measure melody in the style of Bach
  • The Google Doodle was created in honor of prolific composer Bach’s birthday 
  • e-mail

19

View
comments

Google is celebrating composer Johann Sebastian Bach with its first artificial intelligence-powered Doodle.

Thursday’s animated Google Doodle shows the composer playing an organ in celebration of his March 21, 1685, birthday under the old Julian calendar. 

It encourages users to compose their own two-measure melody.

Scroll down for video 


Google is celebrating composer Johann Sebastian Bach with its first artificial intelligence-powered Doodle. Google says the Doodle uses machine learning to ‘harmonize the melody’

Google says the Doodle uses machine learning to ‘harmonize the custom melody into Bach’s signature music style.’  

‘With the press of a button, the Doodle then uses machine learning to harmonize the custom melody into Bach’s signature music style (or a Bach 80’s rock style hybrid if you happen to find a very special easter egg in the Doodle),’ the firm wrote in a blog post.  

Bach’s chorales were known for having four voices carrying their own melodic line.

To develop the AI Doodle, Google teams created a machine-learning model, called Coconet, that was trained on 306 of Bach’s chorale harmonizations. 

  • Canned air, water-spraying drones, and ‘anti-smog tea’:… Facebook Messenger launches new threaded replies to make it… Arctic sea ice continues to dwindle as NASA says its volume… SpaceX delays first ‘hop’ tests of Elon Musk’s Starhopper…

Share this article


The AI Google Doodle was created to celebrate the birthday of prolific composer Johann Sebastian Bach (pictured)

‘His chorales always have four voices, each carrying their own melodic line, while creating a rich harmonic progression when played together,’ the firm said.

‘This concise structure made them good training data for a machine learning model.’  

Google engineers said they trained these machine learning models to spit out ‘polyphonic music in the style of Bach.

‘Coconet is trained to restore Bach’s music from fragments: we take a piece from Bach, randomly erase some notes, and ask the model to guess the missing notes from context,’ according to a blog post on Google’s Magenta unit, which oversees AI projects related to music and art. 

‘The result is a versatile model of counterpoint that accepts arbitrarily incomplete scores as input and works out complete scores.’ 


To develop the AI Doodle, Google teams created a machine-learning model, called Coconet, that was trained on 306 of Bach’s chorale harmonizations

There’s a special feature in the animation that transforms Bach into a 1980s rockstar. Users just click on the mini amplifier next to the keyboard to turn on the synths 

Another team worked to allow machine learning to occur within the web browser instead of on its servers.

The Doodle will prompt users who are unsure of how to interact with the animated graphic.

Users can change they key of the music or its tempo and then share their creation with friends or download it for listening later on. 

Additionally, there’s a special hidden feature in the animation that transforms Bach into a 1980s rockstar.

Users just click on the mini amplifier next to the keyboard to turn on the synths.

HOW DOES ARTIFICIAL INTELLIGENCE LEARN?

AI systems rely on artificial neural networks (ANNs), which try to simulate the way the brain works in order to learn.

ANNs can be trained to recognise patterns in information – including speech, text data, or visual images – and are the basis for a large number of the developments in AI over recent years.

Conventional AI uses input to ‘teach’ an algorithm about a particular subject by feeding it massive amounts of information.   


AI systems rely on artificial neural networks (ANNs), which try to simulate the way the brain works in order to learn. ANNs can be trained to recognise patterns in information – including speech, text data, or visual images

Practical applications include Google’s language translation services, Facebook’s facial recognition software and Snapchat’s image altering live filters.

The process of inputting this data can be extremely time consuming, and is limited to one type of knowledge. 

A new breed of ANNs called Adversarial Neural Networks pits the wits of two AI bots against each other, which allows them to learn from each other. 

This approach is designed to speed up the process of learning, as well as refining the output created by AI systems. 

Source: Read Full Article