Localization and Completion for 3D Object Interactions

Finding where and what objects to put into an existing scene is a common task for scene synthesis and robot/character motion planning. Existing frameworks require development of hand-crafted features suitable for the task, or full volumetric analysis that could be memory intensive and imprecise. In...

Ausführliche Beschreibung

Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on visualization and computer graphics. - 1996. - 26(2020), 8 vom: 14. Aug., Seite 2634-2644
1. Verfasser: Zhao, Xi (VerfasserIn)
Weitere Verfasser: Hu, Ruizhen, Liu, Haisong, Komura, Taku, Yang, Xinyu
Format: Online-Aufsatz
Sprache:English
Veröffentlicht: 2020
Zugriff auf das übergeordnete Werk:IEEE transactions on visualization and computer graphics
Schlagworte:Journal Article
LEADER 01000naa a22002652 4500
001 NLM292687389
003 DE-627
005 20231225073650.0
007 cr uuu---uuuuu
008 231225s2020 xx |||||o 00| ||eng c
024 7 |a 10.1109/TVCG.2019.2892454  |2 doi 
028 5 2 |a pubmed24n0975.xml 
035 |a (DE-627)NLM292687389 
035 |a (NLM)30640616 
040 |a DE-627  |b ger  |c DE-627  |e rakwb 
041 |a eng 
100 1 |a Zhao, Xi  |e verfasserin  |4 aut 
245 1 0 |a Localization and Completion for 3D Object Interactions 
264 1 |c 2020 
336 |a Text  |b txt  |2 rdacontent 
337 |a ƒaComputermedien  |b c  |2 rdamedia 
338 |a ƒa Online-Ressource  |b cr  |2 rdacarrier 
500 |a Date Revised 01.07.2020 
500 |a published: Print-Electronic 
500 |a Citation Status PubMed-not-MEDLINE 
520 |a Finding where and what objects to put into an existing scene is a common task for scene synthesis and robot/character motion planning. Existing frameworks require development of hand-crafted features suitable for the task, or full volumetric analysis that could be memory intensive and imprecise. In this paper, we propose a data-driven framework to discover a suitable location and then place the appropriate objects in a scene. Our approach is inspired by computer vision techniques for localizing objects in images: using an all directional depth image (ADD-image) that encodes the 360-degree field of view from samples in the scene, our system regresses the images to the positions where the new object can be located. Given several candidate areas around the host object in the scene, our system predicts the partner object whose geometry fits well to the host object. Our approach is highly parallel and memory efficient, and is especially suitable for handling interactions between large and small objects. We show examples where the system can hang bags on hooks, fit chairs in front of desks, put objects into shelves, insert flowers into vases, and put hangers onto laundry rack 
650 4 |a Journal Article 
700 1 |a Hu, Ruizhen  |e verfasserin  |4 aut 
700 1 |a Liu, Haisong  |e verfasserin  |4 aut 
700 1 |a Komura, Taku  |e verfasserin  |4 aut 
700 1 |a Yang, Xinyu  |e verfasserin  |4 aut 
773 0 8 |i Enthalten in  |t IEEE transactions on visualization and computer graphics  |d 1996  |g 26(2020), 8 vom: 14. Aug., Seite 2634-2644  |w (DE-627)NLM098269445  |x 1941-0506  |7 nnns 
773 1 8 |g volume:26  |g year:2020  |g number:8  |g day:14  |g month:08  |g pages:2634-2644 
856 4 0 |u http://dx.doi.org/10.1109/TVCG.2019.2892454  |3 Volltext 
912 |a GBV_USEFLAG_A 
912 |a SYSFLAG_A 
912 |a GBV_NLM 
912 |a GBV_ILN_350 
951 |a AR 
952 |d 26  |j 2020  |e 8  |b 14  |c 08  |h 2634-2644