Skip to main content

Continuous Adaptation of Artificial Agents in Complex, Dynamic Environments

Wednesday, March 4, 2020, 11:00 am – 1:00 pm

The proposed research uses advanced AI methods (meta-learning) and cutting-edge technologies including immersive environments and virtual reality (VR) to offer an innovative, productive contribution to the domain of creative storytelling. Specifically, the project aims to propose a solution for problems linked to the creation of dynamic interactive narratives and music compositions, specifically the problem of continuous adaptation of artificial agents in complex dynamic environments. This refers to the difficulty of creating artificial agents that can respond intelligently to evolving complex situations. The main objective of the fellowship is to research and develop new forms of storytelling and composing as a new dimension of artificial intelligence. The project will respond to the challenges by creating a meta-learning framework. In this system, the actor (neural network) learns a high-level procedure (meta-learning) that can be used to generate a good general strategy (policy) in response to environment changes. It is important problem in artificial intelligence, especially in deep learning and it will have many applications in art and music composing.