🇫🇷 Français
🇬🇧 English: I Keep Counting – How did we do it?
🇮🇹️ Italiano

I Keep Counting

The Algomus & Friends team is happy to present I Keep Counting (mp3, webm), our entry for the AI Song Contest 2020 organized by VPRO Broadcast. We finished in 4th place among the 13 competing teams. Congratulations to Uncanney Valley from Australia, and to all the other teams: This event was a nice opportunity to showcase a variety of styles and techniques for using AI in music!

The contest called for European (and Australian, of course) researchers and artists to produce a song inspired by Eurovision through the use of artificial intelligence. We employed a co-creative approach in which the AI and the human fuel each other’s creativity in an iterative process. There is only one human musician that performs in this recording: Niam, who sang I Keep Counting with her warm voice. We bet you wouldn’t have said that!

We hope that you will enjoy listening to the song as much as we did creating it. And… do you want to know more on how we did it?

• 
  

  I stop counting,
  I stop counting,
  It will never end.
  

  I stop counting,
  I stop counting,
  It will never end.

  My heart, my love,
  The world, you know,
  Your eyes, your heart,
  Thank you, my soul...

• One day, the sun,
  The sky, no one,
  I know, at least you come,
  Each day when we go on the war
  between mine and yours,

  I stop counting...

  It's ending, it's ending,
  It's ending, it's ending.

  I keep counting,
  I keep counting,
  I count down the years!

Who we are

We are a research group in Computer Music in the Université de Lille with strong links with the Université de Picardie Jules Verne in Amiens. Following discussions she had at the 2019 ISMIR conference in Delft, Florence Levé set up our AI Song Contest team. Florence has a background in combinatorics and works in Music Information Retrieval, in particular on texture analysis. Louis Bigo focuses on topological music representations and AI methods for music. Richard Groult’s current research is on Music Information Retrieval with a focus on fugue and sonata form analysis. Mathieu Giraud works in computer music, notably on harmony, and in bioinformatics. The team recently expanded, hiring a new member with a different background: Gianluca Micchi, who is betraying his home country Italy to participate for France, where he’s been living since 2013. Funded by the CPER MAuVE, Gianluca is a machine learning researcher coming from physics and his latest project is on the automated functional analysis of harmony. Gianluca wrote a large part of the code for the project, in particular on chord sequence generation and melody generation. Meet him!

Finally, Laurent Feisthauer, a PhD student on cadences, Emmanuel Leguy, an engineer on music interfaces, and David Regnier, a master student on music generation, also contributed to the song. The guest singer, Niam, is a student in engineering and marketing of measurement systems at the Polytech Lille in the Université de Lille.

             

Keep in touch! contact@algomus.fr – @algomus