Quel lien existe-t-il entre la composition automatique de morceaux de musique et l’organisation des musées de demain ? L’intelligence artificielle (IA) investit le champ de la cWhat is the relationship between automatic music composition and the organization of museums of tomorrow? Artificial intelligence (AI) pervades the field of art and questions our creativity. réation artistique et questionne notre rapport à la créativité.
In the last few years, AI art has proliferated. Some AI algorithms can recreate an image in a painter’s style, some can improvise a track alongside a human musician on stage and some can even write a story presenting the reader as hero. Are they as creative or imaginative as humans? Unlike humans, they have a limited degree of functional autonomy and are capable of performing only one function. For example, an AI that’s dedicated to writing will not paint a picture. Mark Riedl, research professor at the Georgia Tech School of Interactive Computing and director of the Entertainment Intelligence Lab throws some light on the issue of AI creativity. When asked by Le Monde, he says: “We are creative when we play Pictionary, when we use a trombone to repair a pair of glasses or when we find another route to go home if a road is closed. Computers already have this kind of creativity.” Jean-Gabriel Ganascia, a researcher at the computer lab at Paris VI, further explains: “Every imagination is seen as the recombination of the elements from pre-existing memory”. This is the underlying principle of how AI creates artistic content.
– Sheherazade IF (for Interactive Fiction) relies on an existing body of text to create its own narrative schemas. It interacts with the reader who gives it the directions to follow and model the books making you the hero.
– EMI and Emily Howell are two AI ‘sisters’: EMI composes music that’s inspired from the work of human composers and also proposes variations. Emily Howell uses EMI’s compositions to produce its own musical compositions.
– Shimon analyzed the improvisational techniques used by pianist Thelonious Monk and it can accompany a live musician with consistent melodies.
Machine learning researchers and artists are also found in the field of imaging. There are painting-inspired IT innovations: Pikazo and Deep Art continue the tradition of pastiche by allowing reproducing any image ‘in the style of’ other. There’s no painter or trend that’s not covered by these apps. Deep Dream, a Google product that aims to better classify images on the basis of their analysis, has even launched a creative stream. This AI can recognize and highlight the human and animal forms in an image. On their blog, their team of engineers has explained the principle of this program. The instruction given is “whatever you see, we want more! If a cloud looks a bit like bird, the [program] will make it look even more like bird.” The psychedelic renderings fascinate web users who love experimenting with this online simulator.
Awarded by Tate Modern, the Italian project Recognition bridges the gap between image recognition and collection management. It screens 1000 photographs a day, supplied by Reuters, and tries to match them with 30,000 artworks in the London museum based on similarities in composition, faces or objects. The ‘matches’ are exhibited in an online gallery, offering an unprecedented perspective on our past and modern representations.
Contrary to the popular fear, AI is not intended to replace artists. Always open to innovation, artists can use AI programs as creative assistants who can accompany them beyond their cognitive abilities.
Predicting the Solar Potential of Rooftops using Image Segmentation and Structured Data How could we automatically estimate the amount of electricity that solar panels could generate on any building of a country, instantly? It would require to be able to […]
How we improved our estimation of vegetated surfaces with Computer Vision Here at nam.R, we’re putting our hard work into building a Digital Twin of the French territory. We aggregate, clean and re-organize a large quantity of data of different […]
In terms of data, tourism industry is complex to analyze. It has just 25% of structured data that comes from varied sources like business websites of tour operators, e-commerce by travel agencies, hosting companies, restaurants, transporters and CRM & other […]
To detect all solar panels on the roofs of french buildings, we used aerial imagery and the known outlines of the buildings, through a pipeline consisting of a solar panel outline detector and a filtering algorithm. We took inspiration from […]
From September 24th to 25th 2018 Charles Hutin-Persillon, Data Strategist and Nicoles Berthelot, Lead Data Strategist, at nam.R participated to the french-norwegian seminar: “Observations, AI and sustainable development” “Observations, AI and sustainable development” Organised by the Ministry of Higher education, […]