BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Melbourne
X-LIC-LOCATION:Australia/Melbourne
BEGIN:DAYLIGHT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
TZNAME:AEDT
DTSTART:19721003T020000
RRULE:FREQ=YEARLY;BYMONTH=4;BYDAY=1SU
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:19721003T020000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260114T163632Z
LOCATION:Darling Harbour Theatre\, Level 2 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231212T093000
DTEND;TZID=Australia/Melbourne:20231212T124500
UID:siggraphasia_SIGGRAPH Asia 2023_sess209_papers_819@linklings.com
SUMMARY:Learning Gradient Fields for Scalable and Generalizable Irregular 
 Packing
DESCRIPTION:Tianyang Xue (Shandong University), Mingdong Wu (Peking Univer
 sity), Lin Lu and Haoxuan Wang (Shandong University), and Hao Dong and Bao
 quan Chen (Peking University)\n\nThe packing problem, also known as cuttin
 g or nesting, has diverse applications in logistics, manufacturing, layout
  design, and atlas generation. It involves arranging irregularly shaped pi
 eces to minimize waste while avoiding overlap. Recent advances in machine 
 learning, particularly reinforcement learning, have shown promise in addre
 ssing the packing problem. In this work, we delve deeper into a novel mach
 ine learning-based approach that formulates the packing problem as conditi
 onal generative modeling. To tackle the challenges of irregular packing, i
 ncluding object validity constraints and collision avoidance, our method e
 mploys the score-based diffusion model to learn a series of gradient field
 s. These gradient fields encode the correlations between constraint satisf
 action and the spatial relationships of polygons, learned from teacher exa
 mples. During the testing phase, packing solutions are generated using a c
 oarse-to-fine refinement mechanism guided by the learned gradient fields. 
 To enhance packing feasibility and optimality, we introduce two key archit
 ectural designs: multi-scale feature extraction and coarse-to-fine relatio
 n extraction. We conduct experiments on two typical industrial packing dom
 ains, considering translations only. Empirically, our approach demonstrate
 s spatial utilization rates comparable to, or even surpassing, those achie
 ved by the teacher algorithm responsible for training data generation. Add
 itionally, it exhibits some level of generalization to shape variations. W
 e are hopeful that this method could pave the way for new possibilities in
  solving the packing problem.\n\nRegistration Category: Full Access, Enhan
 ced Access, Trade Exhibitor, Experience Hall Exhibitor\n\n
URL:https://asia.siggraph.org/2023/full-program?id=papers_819&sess=sess209
END:VEVENT
END:VCALENDAR