Ai Composing Emotional Soundtrack / Music Compositions Artificial intelligence
Artificial intelligence (AI) and research into music compositions began more than a decade ago, and several organizations have developed AI technologies to compose music without human intervention. However, a new wave has emerged that has the potential for artificial intelligence in the creative economy, including musical composition. Startups and incumbents are using AI to make music, writing text and some have been very successful.
Aiva Technologies was established in 2016 and its team is working on developing AI scripts to create emotional soundtracks for ads, videos, games, and movies. AI music composers create original, copyright-free music that you can use in your latest YouTube videos, social media videos, and ads. You need to be a sound designer or musician to produce a soundtrack for your videos. AI music composers allow you to upload pre-recorded music and create variations of it.
It acts as a creative assistant to help songwriters with their music production. It can also be used to make variations on existing songs by allowing its users to create music from scratch. Aiva Technologies released its first title, Aopus 1: Piano Soloa, and in subsequent years released albums and composed music for video games.
As a virtual artist of artificial intelligence, Aiva has been trained in over 30,000 classical compositions and relies on mathematical algorithms to represent patterns from music theory. Aiva allows its users to define metrics such as genre, and it lists twelve genres available to its users, along with tempo and length.
Aiva composes picture-synchronous music based on visual information from video games. In real-time synchronized composition data such as game intensity, colors and actions are interpreted and the soundtrack is composed to match the events on the screen.
In a recent demonstration conducted by Zoom Watch, the music composition app Dynascore reformed the entire emotional tenor of a short video several times per minute without changing a single frame. Starting with a brief briefing and a serious training display with a serious soundtrack, you would expect neon sweat to pour out of an athlete’s head and the Gatorade logo to change the tone to something funnier. But using the machine’s intelligence engine, Dynasc swapped the action film score for Beethoven’s gloomy Moonlight Sonata and turned the video into a dark comedy.
Unlike many film scores, where a single theme is repeated several times in a soundtrack, we wanted to apply theme variations to our Chinese album, where repetition of a theme is presented in various tracks. To achieve this, we used the techniques of deep learning to create musical variations on the theme of the album. Our main aim for further work is to develop music generators for music therapy that can produce music that evokes certain emotions in the listener. The approach described here is suitable for the development of generic music generators capable of generating specific emotions in the audience, while functional music with non-linear duration would be useful for video games, soundtracking, and creative industries, as interpreted here.
This work suggests that generative music technology has the potential to generate infinite soundtracks of sympathy for the listener via biosignals and biofeedback loops. Promising treatments combine music with emotions in the creative industries, art therapy, and relaxation.
Pierre Barreau is co-founder and CEO of Aiva, artificial intelligence that composes emotional soundtrack music and is one of the most innovative minds at the interface of art and technology. Aiva Technologies was founded in Luxembourg and London last year by Pierre Barreau, Denis Shtefan, Arnaud Decker, and Vincent Barreau. They created an artificial intelligence called AIVA (Artificial Intelligence Virtual Artist) and learned her how to compose classical music and emotional art that is considered human.
Aiva learns by capturing concepts from music theory and understanding the art of music composition by reading large databases of classical partitions by famous composers. These partitions are used to train Aiva until the copyright expires, allowing the AI to educate itself for free. The first generative music software, Koan, was released by Sseyo in 1994.
The AIVA team decided to focus on classical music because it is a style of music used in films, games, commercials, trailers, and soundtracks. Thanks to recent developments in AI technology and machine learning, which allow original and personal music to be composed, creators can offer a solution that consumes fewer resources and less time. However, they still work with copyright restrictions and compositions that lack originality and adaptability.
