Output details
11 - Computer Science and Informatics
Lancaster University
Modeling spatial and temporal variation in motion data
<26> Proposes a fundamentally different approach to the problem of synthesising motion variation, featuring two important ideas: we are the first to take a data-driven approach to solve this problem, and the first to use Dynamic Bayesian Networks to model motion variation. Published in ACM Transactions on Graphics (world-leading journal in the area), and presented at the annual SIGGRAPH Asia conference (acceptance ratio: 70/275). Collaboration with faculty from Carnegie Mellon University. High citation rate is evidence for follow-up work. Using Bayesian Networks for human motion analysis has since led to significant research funding in the US by other researchers.