Download:
|
by Mark F. Westling, Larry S. Davis
Lecture Notes in Computer Science
http://www.umiacs.umd.edu/users/westling/Papers/accv98.ps.gz
Add To MetaCart
Abstract:
Abstract. In most object recognition systems, interactions between objects in a scene are ignored and the best interpretation is considered to be the set of hypothesized objects that matches the greatest number of image features. We show how image interpretation can be cast as the problem of finding the most probable explanation (MPE) in a Bayesian network that models both visual and physical object interactions. The problem of how to determine exact conditional probabilities for the network is shown to be unimportant, since the goal is to find the most probable configuration of objects, not to calculate absolute probabilities. We furthermore show that evaluating configurations by feature counting is equivalent to calculating the joint probability of the configuration using a restricted Bayesian network, and derive the assumptions about probabilities necessary to make a Bayesian formulation reasonable. 1
Citations
|
635
|
An Introduction to Bayesian Networks
– Jensen
- 1996
|
|
308
|
Object recognition by computer: the role of geometric constraints
– Grimson
- 1990
|
|
217
|
Recognizing Solid Objects by Alignment with an Image
– Huttenlocher, Ullman
- 1990
|
|
77
|
Control of Selective Perception Using Bayes Nets and Decision Theory
– Rimey, Brown
- 1994
|
|
54
|
Fast Recognition using Adaptive Subdivisions of Transformation Space
– Breuel
- 1992
|
|
31
|
Model-Based influence diagrams for machine vision
– Levitt, Agosta, et al.
- 1990
|
|
30
|
Bayesian inference in model-based machine vision
– Binford, Levitt, et al.
- 1989
|
|
29
|
Polynomial-Time Object Recognition in the Presence of Clutter, Occlusion, and Uncertainty
– Cass
- 1992
|
|
27
|
Local Search Algorithms for Geometric Object Recognition: Optimal Correspondence and Pose
– Beveridge
- 1993
|
|
26
|
Distributed bayesian object recognition
– Rigoutsos, Hummel
- 1993
|
|
22
|
Efficient inference in Bayes nets as a combinatorial optimization problem
– Li, D'Ambrosio
- 1994
|
|
21
|
CAGD-Based Computer Vision
– Hansen, Henderson
- 1989
|
|
16
|
Image interpretation using Bayesian networks
– Kumar, Desai
- 1996
|
|
12
|
Utility-based control for computer vision
– Levitt, Binford, et al.
- 1990
|
|
8
|
Higher-Order Statistics in Object Recognition
– Breuel
- 1993
|
|
6
|
Model-based recognition of objects in complex scenes
– Binford, Levitt
- 1994
|
|
6
|
Visual surveillance in a dynamic and uncertain
– Buxton, Gong
- 1995
|
|
6
|
Reasoning about occlusions during hypothesis verification
– Rothwell
- 1996
|
|
5
|
SUCCESSOR: Interpretation Overview and Constraint System
– Mann, Binford
- 1996
|
|
5
|
Probabilistic Resoning in Intelligent Systems: Networks of Plausible Inference
– Pearl
- 1988
|
|
4
|
Distributed Belief Revision for Adaptive Image Processing Regulation
– Murino, Peri, et al.
- 1992
|
|
3
|
A framework for generic object recognition with bayesian networks
– Liang, Jensen, et al.
- 1996
|
|
3
|
Efficient Enumeration of Instantiations in Bayesian Networks
– Srinivas, Nayak
- 1996
|
|
2
|
Object recognition by fast hypothesis generation and reasoning about object interactions
– Westling, Davis
- 1996
|
|
1
|
SPI in large BN2O networks
– D'Ambrosio
- 1994
|
|
1
|
Performance comparison of ten variations on the interpretation tree matching algorithm
– Fisher
- 1994
|
|
1
|
Statistical Pattern Recognition
– Wells
- 1993
|
