Search Results

< Back Task Relabelling for Multi-task Transfer using Successor Features Link Author(s) M Balla, D Perez-Liebana Abstract More info TBA Link

< Back Emergence in the Expressive Machine Link Author(s) L Dekker Abstract More info TBA Link

< Back Nice is Different than Good: Longitudinal Communicative Effects of Realistic and Cartoon Avatars in Real Mixed Reality Work Meetings Link Author(s) GC Dobre, M Wilczkowiak, M Gillies, X Pan, S Rintel Abstract More info TBA Link

< Back Comparative evaluation in the wild: Systems for the expressive rendering of music Link Author(s) K Worrall, Z Yin, T Collins Abstract More info TBA Link

< Back A Qualitative Investigation of Real World Accessible Design Experiences within a Large Scale Commercial Game Development Studio Link Author(s) J Kulik, P Cairns Abstract More info TBA Link

< Back How does machine learning affect diversity in evolutionary search? Procedural content generation of video games levels has greatly benefited from machine learning. In such complex domains, generative models can provide representation spaces for evolutionary search. But how expressive are such learned models? How many different levels would they be able to produce? A new paper, co-authored by IGGI PhD researcher Sebastian Berns and Professor Simon Colton, looks at the limitations of generative models in the context of multi-solution optimisation. The work will be presented at the Genetic and Evolutionary Computation Conference (GECCO) and is nominated for a best paper award . The study shows that quality diversity (QD) search in the latent space of a variational auto-encoder yields a solution set of lower diversity than in a manually-defined genetic parameter space. The authors find that learned latent spaces are useful for the comparison of artefacts and recommend their use for distance and similarity estimation. However, whenever a parametric search space is obtainable, it should be preferred over a learned representation space as it produces a higher diversity of solutions. Alexander Hagg, Sebastian Berns, Alexander Asteroth, Simon Colton & Thomas Bäck. (2021). Expressivity of Parameterized and Data-driven Representations in Quality Diversity Search. In Proceedings of the Genetic and Evolutionary Computation Conference. Pre-print available on arXiv https://arxiv.org/abs/2105.04247Accompanying code repository available on Github https://github.com/alexander-hagg/ExpressivityGECCO2021 Previous 27 Jun 2021 Next

< Back iGGi Studentships - Home Fees Candidates iGGi SPECIAL RECRUITMENT ROUND - home-fees only - With focus on home candidates only, we are inviting applications for studentships for the iGGi PhD Programme. Each studentship includes four years of fully funded (fees and stipend at UKRI rate) full-time study starting September 2023. The PhD researchers will be based at Queen Mary University of London or University of York (depending on which uni the chosen primary supervisor belongs to). To apply please follow the instructions on our Apply page Submit your full application by Monday 15 May 2023, 12:00 noon BST. Previous 5 Apr 2023 Next

< Back Expressive curve editing with the sigma lognormal model Link Author(s) D Berio, FF Leymarie, R Plamondon Abstract More info TBA Link

< Back Monte carlo tree search applied to co-operative problems Link Author(s) PR Williams, J Walton-Rivers, D Perez-Liebana, SM Lucas Abstract More info TBA Link

< Back System and method for training a machine learning model Link Author(s) R Spick, G Moss, T Bradley, PV Amadori Abstract More info TBA Link

< Back Tackling sparse rewards in real-time games with statistical forward planning methods Link Author(s) RD Gaina, SM Lucas, D Perez-Liebana Abstract More info TBA Link

< Back NoiseBandNet: controllable time-varying neural synthesis of sound effects using filterbanks Link Author(s) A Barahona-Rios, T Collins Abstract More info TBA Link