TY - GEN
T1 - Divided we fall
T2 - 3rd European Conference on Computer Vision, ECCV 1994
AU - Cooper, Paul R.
AU - Birnbaum, Lawrence A
AU - Halabe, Daniel
AU - Brand, Matthew
AU - Prokopowicz, Peter N.
N1 - Funding Information:
Many thanks are due to R. Tournebize for critical reading the manuscript. Special thanks are extended to P. Sansonetti for his support to our research project and to P. Roux for his advice in confocal microscopy analysis. This work was supported by grants from the French Ministère de l'Education Nationale, de l'Enseignement Supérieur, de la Recherche et de l'Insertion Professionnelle, and from the NORD-SUD INSERM program (Grant No. 4N0016). The confocal microscope used in this work is a gift of M. and L. Pollack.
Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 1994.
PY - 1994
Y1 - 1994
N2 - An image of a scene with occlusions can yield only partial knowledge about disconnected fragments of the scene. If this were the only knowledge available, programs attempting to interpret the scene would have to conclude that the scene fragments would collapse in a jumble. But they won't. We describe a program that exploits commonsense knowledge of naive physics to make sense of scenes with occlusion. Our causal analysis focuses on the static stability of structures: what supports what. Occluded connections in a link-and-junction scene are inferred by determining the stability of each subassembly in the scene, and connecting parts when they are unstable. The causal explanation that is generated reflects a deeper understanding of the scene than mere model matching; it allows the seeing agent to predict what will happen next in the scene, and determine how to interact with it.
AB - An image of a scene with occlusions can yield only partial knowledge about disconnected fragments of the scene. If this were the only knowledge available, programs attempting to interpret the scene would have to conclude that the scene fragments would collapse in a jumble. But they won't. We describe a program that exploits commonsense knowledge of naive physics to make sense of scenes with occlusion. Our causal analysis focuses on the static stability of structures: what supports what. Occluded connections in a link-and-junction scene are inferred by determining the stability of each subassembly in the scene, and connecting parts when they are unstable. The causal explanation that is generated reflects a deeper understanding of the scene than mere model matching; it allows the seeing agent to predict what will happen next in the scene, and determine how to interact with it.
UR - http://www.scopus.com/inward/record.url?scp=85027603124&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85027603124&partnerID=8YFLogxK
U2 - 10.1007/3-540-57956-7_61
DO - 10.1007/3-540-57956-7_61
M3 - Conference contribution
AN - SCOPUS:85027603124
SN - 9783540579564
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 535
EP - 540
BT - Computer Vision — ECCV 1994 - 3rd European Conference on Computer Vision, Proceedings
A2 - Eklundh, Jan-Olof
PB - Springer Verlag
Y2 - 2 May 1994 through 6 May 1994
ER -