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:20260817T171531Z
LOCATION:Hall B7 (1)\, B Block\, Level 7
DTSTART;TZID=Asia/Tokyo:20241204T110800
DTEND;TZID=Asia/Tokyo:20241204T111900
UID:siggraphasia_SIGGRAPH Asia 2024_sess114_papers_508@linklings.com
SUMMARY:CPoser: An Optimization-after-Parsing Approach for Text-to-Pose Ge
 neration Using Large Language Models.
DESCRIPTION:Yumeng Li, Bohong Chen, Zhong Ren, and Yao-Xiang Ding (Zhejian
 g University); Libin Liu (Peking University); and Tianjia Shao and Kun Zho
 u (Zhejiang University)\n\nText-to-pose generation is challenging due to t
 he complexity of natural language and human posture semantics. Utilizing l
 arge language models (LLMs) for text-to-pose generation is appealing due t
 o their strong capabilities in text understanding and reasoning. However, 
 as LLMs are designed for general-purpose language processing and not speci
 fically trained for pose generation, it remains nontrivial to generate pre
 cise articulation targets for the full body using LLMs directly. To this e
 nd, we propose CPoser, a novel approach to harness the power of LLMs for t
 ext-to-pose generation, featuring a prompt parsing stage and a pose optimi
 zation stage. The parsing stage utilizes LLMs to turn text prompts into po
 se intermediate representations (Pose-IRs) through a set of predefined str
 uctured queries. These Pose-IRs explicitly describe specific pose conditio
 ns, such as squatting depth and knee bending angle, naturally forming an o
 bjective function that a target pose should satisfy. The optimization stag
 e solves for expressive poses and hand gestures based on the Pose-IR objec
 tive function via robust optimization in a quantized pose prior space. The
  results are further refined to enhance naturalness and incorporate facial
  expressions. Experiments show that our approach effectively understands d
 iverse text prompts for pose generation, surpassing existing text-to-pose 
 methods.\n\nRegistration Category: Full Access, Full Access Supporter\n\nL
 anguage Format: English Language\n\nSession Chair: Kai Wang (Simon Fraser 
 University, ShanghaiTech University)\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_508&sess=sess114
END:VEVENT
END:VCALENDAR