BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Tokyo
X-LIC-LOCATION:Asia/Tokyo
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:JST
DTSTART:18871231T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260817T171534Z
LOCATION:Hall C\, C Block\, Level 4
DTSTART;TZID=Asia/Tokyo:20241206T160000
DTEND;TZID=Asia/Tokyo:20241206T180000
UID:siggraphasia_SIGGRAPH Asia 2024_sess208_real_118@linklings.com
SUMMARY:Prototyping Game Character Animations with AI in Unity
DESCRIPTION:Florent Bocquelet, Félix Harvey, and Pierre-Luc Loyer (Unity T
 echnologies)\n\nMost 3D games leverage animations to breathe life into cha
 racters. Even at the early stages of game development, prototyping with wo
 rking animations helps developers quickly understand and adapt the game de
 sign. Controllable characters require animations that are constrained in m
 ultiple ways in order to blend together and avoid artifacts such as feet s
 liding. Authoring, modifying and iterating on such animations is a costly 
 process that often requires high level expertise or expensive motion captu
 re.\n\nWe present a novel AI-powered graph tool that allows users to lever
 age generative machine learning models, presented as nodes in the graph, a
 nd create animations from only a few high-level inputs. This greatly reduc
 es the time and expertise required to author animations from scratch, whil
 e offering unseen levels of control with generative motion creation.\n\nTh
 e combination of the many relevant nodes that the graph supports and the f
 lexibility of our state-of-the-art diffusion models allows for going beyon
 d text-to-motion and in-betweening, common in the literature. We can gener
 ate motions that can be synchronized on multiple aspects, such as duration
 , speed, feet contact timings, while being constrained by signals such as 
 text-prompts and trajectories. Our diffusion models also allow for stylizi
 ng existing motions, and generating looping animations. This makes it poss
 ible, for the first time to our knowledge, to generate complete working lo
 comotion sets from high level inputs such as text and duration, directly u
 sable with common character controllers.\n\nOur system being implemented i
 n Unity, we showcase the power of this novel approach by constructing a pr
 ototype animated character with working locomotion and attack motions with
 out pre-existing animations, in a few minutes.\n\nRegistration Category: E
 nhanced Access, Full Access, Full Access Supporter\n\nLanguage Format: Eng
 lish Language\n\nSession Chair: Takahito Tejima (Polyphony Digital Inc.)\n
 \n
URL:https://asia.siggraph.org/2024/program/?id=real_118&sess=sess208
END:VEVENT
END:VCALENDAR