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:G502\, G Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241205T130000
DTEND;TZID=Asia/Tokyo:20241205T132000
UID:siggraphasia_SIGGRAPH Asia 2024_sess295_edu_121@linklings.com
SUMMARY:An Eye for an AI: Evaluating GPT-4o's Visual Perception Skills and
  Geometric Reasoning Skills Using Computer Graphics Questions
DESCRIPTION:Tony Haoran Feng, Paul Denny, Burkhard C. Wünsche, and Andrew 
 Luxton-Reilly (University of Auckland) and Jacqueline Whalley (Auckland Un
 iversity of Technology)\n\nCG (Computer Graphics) is a popular field of CS
  (Computer Science), but many students find this topic difficult due to it
  requiring a large number of skills, such as mathematics, programming, geo
 metric reasoning, and creativity. Over the past few years, researchers hav
 e investigated ways to harness the power of GenAI (Generative Artificial I
 ntelligence) to improve teaching. In CS, much of the research has focused 
 on introductory computing. A recent study evaluating the performance of an
  LLM (Large Language Model), GPT-4 (text-only), on CG questions, indicated
  poor performance and reliance on detailed descriptions of image content, 
 which often required considerable insight from the user to return reasonab
 le results. So far, no studies have investigated the abilities of LMMs (La
 rge Multimodal Models), or multimodal LLMs, to solve CG questions and how 
 these abilities can be used to improve teaching.\n\nIn this study, we cons
 truct two datasets of CG questions requiring varying degrees of visual per
 ception skills and geometric reasoning skills, and evaluate the current st
 ate-of-the-art LMM, GPT-4o, on the two datasets. We find that although GPT
 -4o exhibits great potential in solving questions with visual information 
 independently, major limitations still exist to the accuracy and quality o
 f the generated results. We propose several novel approaches for CG educat
 ors to incorporate GenAI into CG teaching despite these limitations. We ho
 pe that our guidelines further encourage learning and engagement in CG cla
 ssrooms.\n\nRegistration Category: Enhanced Access, Full Access, Full Acce
 ss Supporter\n\nLanguage Format: English Language\n\nSession Chair: xuejun
  xuan (China Academy of Art, School of Animation and Games)\n\n
URL:https://asia.siggraph.org/2024/program/?id=edu_121&sess=sess295
END:VEVENT
END:VCALENDAR