DocumentCode :
3312774
Title :
Overcomplete Dictionary Design: The Impact of the Sparse Representation Algorithm
Author :
Irofti, Paul ; Dumitrescu, Bogdan
Author_Institution :
Dept. of Autom. Control & Comput., Univ. Politeh. of Bucharest, Bucharest, Romania
fYear :
2015
fDate :
27-29 May 2015
Firstpage :
901
Lastpage :
908
Abstract :
The design of dictionaries for sparse representations is typically done by iterating two stages: compute sparse representations for the fixed dictionary and update the dictionary using the fixed representations. Most of the innovation in recent work was proposed for the update stage, while the representation stage was routinely done with Orthogonal Matching Pursuit (OMP), due to its low complexity. We investigate here the use of other greedy sparse representation algorithms, more computationally demanding than OMP but still with convenient complexity. These algorithms include a new proposal, the projection-based Orthogonal Least Squares. It turns out that the effect of using better representation algorithms may be more significant than improving the update stage, sometimes even leveling the performance of different update algorithms. The numerous experimental results presented here suggest which are the best combinations of methods and open new ways of designing and using dictionaries for sparse representations.
Keywords :
learning (artificial intelligence); signal representation; OMP; fixed representation; greedy sparse representation algorithm; orthogonal matching pursuit; overcomplete dictionary design; projection-based orthogonal least squares; sparse representation algorithm; update algorithms; Algorithm design and analysis; Complexity theory; Correlation; Dictionaries; Matching pursuit algorithms; Signal processing algorithms; Sparse matrices; dictionary learning; greedy algorithms; sparse representation;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Control Systems and Computer Science (CSCS), 2015 20th International Conference on
Conference_Location :
Bucharest
Print_ISBN :
978-1-4799-1779-2
Type :
conf
DOI :
10.1109/CSCS.2015.50
Filename :
7168533
Link To Document :
بازگشت