Title :
The Tower of Knowledge Scheme for Learning in Computer Vision
Author :
Petrou, Maria ; Xu, Mai
Abstract :
A scheme, named tower of knowledge (ToK), is proposed for interpreting 3D scenes. The ToK encapsulates causal dependencies between object appearance and functionality. We demonstrate it by labelling the components of the 3D model of a building, reconstructed from images of multiple views, by using utility theory.
Keywords :
Bayesian methods; Computer vision; Digital images; Fires; Humans; Layout; Object recognition; Ovens; Poles and towers; Utility theory;
Conference_Titel :
Digital Image Computing Techniques and Applications, 9th Biennial Conference of the Australian Pattern Recognition Society on
Conference_Location :
Glenelg, Australia
Print_ISBN :
0-7695-3067-2
DOI :
10.1109/DICTA.2007.4426780
