Luke Farrar

University of York

Luke Farrar is an iGGi PhD student at The University of York undertaking research in Flexible and Realistic Character Animations in Complex and Dynamic Environments. Luke's research focuses through his bachelor's and master's degrees were on applying machine learning to interesting and unique settings. In his bachelor's he focused on creating an application for individuals that suffered from cognitive impairments through the use of the "Microsoft HoloLens" and machine learning to allow those individuals to maintain a semblance of everyday life. In his postgraduate Luke focused on using machine learning to generalise high-fidelity scientific simulations to rapidly generate predictions for parameter combinations that had not yet been sampled in order to accelerate the production of new results. Luke revels in all things AI, knowing that there is always more to learn and seeks to continually deepen his understanding around AI.

A description of Luke's research:

Modern games have an increasing focus on hyper-realism and immersion to better capture the attention of players. One of the ways that games can break this immersion is by having animations that break the flow of movement or actions through the use of predefined animations. Motion matching is a solution for predicting the best next frame of an animation by looking at the pose and user trajectory. The downside however, is that when you increase the amount of possible animations in the database the runtime cost also increases. A solution was proposed known as 'learned motion matching' (Holden et al., 2020) which takes the positive properties of motion matching but also achieves the scalability of neural-network-based generative models. This project will explore and improve the learned motion matching method through implementation of memory layers to improve accuracy without the sacrifice of increasing runtime costs. A restructuring and adaptation of the existing machine learning neural network used could also improve the learned motion matching method as breaking down each step of the learned motion matching at each step could uncover optimisations that are not initially visible. Another way restructuring could improve the learned motion matching is through creating a more succinct all-in-one approach which may streamline the process.