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:20250110T023302Z
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:Educator's Forum\n\nTony Haoran Feng, Paul Denny, Burkhard C. 
 Wünsche, and Andrew Luxton-Reilly (University of Auckland) and Jacqueline 
 Whalley (Auckland University 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 mathemati
 cs, programming, geometric reasoning, and creativity. Over the past few ye
 ars, researchers have investigated ways to harness the power of GenAI (Gen
 erative Artificial Intelligence) to improve teaching. In CS, much of the r
 esearch has focused on introductory computing. A recent study evaluating t
 he performance of an LLM (Large Language Model), GPT-4 (text-only), on CG 
 questions, indicated poor performance and reliance on detailed description
 s of image content, which often required considerable insight from the use
 r to return reasonable results. So far, no studies have investigated the a
 bilities of LMMs (Large Multimodal Models), or multimodal LLMs, to solve C
 G questions and how these abilities can be used to improve teaching.\n\nIn
  this study, we construct two datasets of CG questions requiring varying d
 egrees of visual perception skills and geometric reasoning skills, and eva
 luate the current state-of-the-art LMM, GPT-4o, on the two datasets. We fi
 nd that although GPT-4o exhibits great potential in solving questions with
  visual information independently, major limitations still exist to the ac
 curacy and quality of the generated results. We propose several novel appr
 oaches for CG educators to incorporate GenAI into CG teaching despite thes
 e limitations. We hope that our guidelines further encourage learning and 
 engagement in CG classrooms.\n\nRegistration Category: Enhanced Access, Fu
 ll Access, Full Access Supporter\n\nLanguage Format: English Language\n\nS
 ession Chair: xuejun xuan (China Academy of Art, School of Animation and G
 ames)
URL:https://asia.siggraph.org/2024/program/?id=edu_121&sess=sess295
END:VEVENT
END:VCALENDAR