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:20240214T070241Z
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_419@linklings.com
SUMMARY:Neural Packing: from Visual Sensing to Reinforcement Learning
DESCRIPTION:Technical Papers\n\nJuzhan Xu (Shenzhen University), Minglun G
 ong (University of Guelph), Hao Zhang (Simon Fraser University), and Hui H
 uang and Ruizhen Hu (Shenzhen University)\n\nWe present a novel learning f
 ramework to solve the transport-and-packing (TAP) problem in 3D. It consti
 tutes a full solution pipeline from partial observations of input objects 
 via RGBD sensing and recognition to final box placement, via robotic motio
 n planning, to arrive at a compact packing in a target container. The tech
 nical core of our method is a neural network for TAP, trained via reinforc
 ement learning (RL), to solve the NP-hard combinatorial optimization probl
 em. Our network simultaneously selects an object to pack and determines th
 e final packing location, based on a judicious encoding of the continuousl
 y evolving states of partially observed source objects and available space
 s in the target container, using separate encoders both enabled with atten
 tion mechanisms. The encoded feature vectors are employed to compute the m
 atching scores and feasibility masks of different pairings of box selectio
 n and available space configuration for packing strategy optimization. Ext
 ensive experiments, including ablation studies and physical packing execut
 ion by a real robot (Universal Robot UR5e), are conducted to evaluate our 
 method in terms of its design choices, scalability, generalizability, and 
 comparisons to baselines, including the most recent RL-based TAP solution.
  We also contribute the first benchmark for TAP which covers a variety of 
 input settings and difficulty levels.\n\nRegistration Category: Full Acces
 s, Enhanced Access, Trade Exhibitor, Experience Hall Exhibitor
URL:https://asia.siggraph.org/2023/full-program?id=papers_419&sess=sess209
END:VEVENT
END:VCALENDAR