Pleased to announce the ICRA 2021 proceedings are now available, including: Swarm Map-based Bayesian Optimisation for allowing centralised and decentralised robot swarms to adapt rapidly to changes in the environment https://ieeexplore.ieee.org/document/9560958 ASVLite, a high-performance simulator for autonomous surface vehicles based on frequency domain analysis https://ieeexplore.ieee.org/document/9561815
Want to build a repertoire of solutions that generalises well to new problems? Or customise a repertoire of solutions to any kind of meta-objective? QD-Meta might be your best bet. QD-Meta, shorthand for Quality-Diversity Meta-Evolution, evolves a population of candidate repertoires by optimising the behaviour space to a user-defined meta-objective. Read the new publication here.