Paper Info
Title
Architecture Design of Belief Propagation for Real-Time Disparity Estimation
Abstract
Belief propagation based algorithms perform best in disparity estimation but suffer from high computational complexity and storage, especially in message passing. This paper proposes an efficient architecture design with three techniques to solve the problems. For the memory storage, we propose the spinning-message and the sliding-bipartite node plane that can reduce memory cost to 1.2% for image-scale algorithms and 23.4% for block-scale algorithms, when compared to the traditional approach. For the logic complexity, we propose a buffer-free processing element architecture that has 3.6 times hardware efficiency of the previous work. The three proposed techniques could be applied to various belief propagation based algorithms to save significant hardware cost as well as approach real-time speed.
Year
DOI
Venue
2010
10.1109/TCSVT.2010.2087434
IEEE Trans. Circuits Syst. Video Techn.
Keywords
Field
DocType
times hardware efficiency,belief networks,belief propagation,image-scale algorithm,approach real-time speed,efficient architecture design,storage management,logic complexity,block-scale algorithm,real-time disparity estimation,significant hardware cost,high computational complexity,belief maintenance,buffer-free processing element architecture,graph theory,memory cost,message passing,architecture design,spinning-message,memory storage,sliding-bipartite node plane,disparity estimation,algorithm design and analysis,estimation,memory management,computational complexity,real time systems,real time
Computer science,Real-time computing,Memory management,Artificial intelligence,Computer engineering,Integrated circuit,Message passing,Belief propagation,Graph theory,Computer vision,Architecture,Algorithm design,Computational complexity theory
Journal
Volume
Issue
ISSN
20
11
1051-8215
Citations 
PageRank 
References 
2
0.41
22
Authors
2
Name
Order
Citations
PageRank
Yu-cheng Tseng11068.90
Tian-Sheuan Chang271269.10