Meta-learning for acoustic and music composing

• Published in: The Journal of the Acoustical Society of America 2020-10, Vol.148 (4), p.2701-2701
• Main Author: Lisek, Robert B.
• Format: Article
• Language: English
• ISSN: 0001-4966; 1520-8524
• DOI: 10.1121/1.5147483

We observe the success of artificial neural networks in simulating human performance on a number of tasks: such as image or sound recognition, natural language processing, etc. However, there are limits to state-of-the- art AI that separate it from human-like intelligence. Humans can learn a new skill without forgetting what they have already learned and they can improve their activity and gradually become better learners. Today's AI algorithms are limited in how much previous knowledge they are able to keep through each new training phase and how much they can reuse. In practice this means that it is necessary to build and adjust new algorithms to every new particular task. This is closer to a sophisticated data processing than to real intelligence. This is why research concerning generalisation are becoming increasingly important. A generalization in AI means that system can generate new compositions or find solutions for new tasks that are not present in the training corpus. Intelligent agent should have meta learning capabilities, should not just be able to memorize the solution to a fixed set of tasks during creating of stories, but learn how to generalize to new problems it encounters. The project proposes a solution for problems linked to organizing acoustic material and creation of new compositions based on meta-learning. Meta-Composer is a neural network equipped with the ability to combine acoustic materials and partial compositions in a flexible and combinatorial way to create a new consistent general composition.