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:20260817T171530Z
LOCATION:Hall B7 (1)\, B Block\, Level 7
DTSTART;TZID=Asia/Tokyo:20241206T131400
DTEND;TZID=Asia/Tokyo:20241206T132800
UID:siggraphasia_SIGGRAPH Asia 2024_sess147_papers_977@linklings.com
SUMMARY:SIGGesture: Generalized Co-Speech Gesture Synthesis via Semantic I
 njection with Large-Scale Pre-Training Diffusion Models
DESCRIPTION:Qingrong Cheng (Tencent AI Lab, Tencent TIMI L1 Studio) and Xu
  Li and Xinghui Fu (Tencent AI Lab)\n\nThe automated synthesis of high-qua
 lity 3D gestures from speech holds significant value for virtual humans an
 d gaming. Previous methods primarily focus on synchronizing gestures with 
 speech rhythm, often neglecting semantic gestures. These semantic gestures
  are sparse and follow a long-tailed distribution across the gesture seque
 nce, making them challenging to learn in an end-to-end manner. Additionall
 y, generating rhythmically aligned gestures that generalize well to in-the
 -wild speech remains a significant challenge. To address these issues, we 
 introduce SIGGesture, a novel diffusion-based approach for synthesizing re
 alistic gestures that are both high-quality and semantically pertinent. Sp
 ecifically, we firstly build a robust diffusion-based foundation model for
  rhythmical gesture synthesis by pre-training it on a collected large-scal
 e dataset with pseudo labels. Secondly,  we leverage the powerful generali
 zation capabilities of Large Language Models (LLMs) to generate appropriat
 e semantic gestures for various speech transcripts. Finally, we propose a 
 semantic injection module to infuse semantic information into the synthesi
 zed results during the diffusion reverse process. Extensive experiments de
 monstrate that SIGGesture significantly outperforms existing baselines, ex
 hibiting excellent generalization and controllability.\n\nRegistration Cat
 egory: Full Access, Full Access Supporter\n\nLanguage Format: English Lang
 uage\n\nSession Chair: Yi Zhou (Roblox)\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_977&sess=sess147
END:VEVENT
END:VCALENDAR