Abstract
The performance of sparse recovery using compressed measurements improves when dictionaries learned from training data are used in place of predefined dictionaries. In this paper, we propose to learn incoherent multilevel dictionaries using discriminative clustering in each level. To this end, we present the discriminative K-lines clustering that iterates between identifying the cluster centers and computing the discriminant directions. A scheme for computing representations using the proposed dictionary is also developed. Simulation results for compressed sensing using standard images demonstrate that incorporating incoherence in the dictionary results in improved recovery performance. Furthermore, we implement the proposed algorithms as part of a sparse representations toolbox for the J-DSP software package.
| Original language | English (US) |
|---|---|
| Title of host publication | Proceedings of the 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012 |
| Pages | 494-497 |
| Number of pages | 4 |
| DOIs | |
| State | Published - 2012 |
| Event | 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012 - Piraeus-Athens, Greece Duration: Jul 18 2012 → Jul 20 2012 |
Other
| Other | 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012 |
|---|---|
| Country | Greece |
| City | Piraeus-Athens |
| Period | 7/18/12 → 7/20/12 |
Fingerprint
Keywords
- compressed sensing
- Discriminative clustering
- incoherent dictionaries
- sparse representations
ASJC Scopus subject areas
- Computer Graphics and Computer-Aided Design
- Computer Vision and Pattern Recognition
- Signal Processing
Cite this
Learning multilevel dictionaries for compressed sensing using discriminative clustering. / Thiagarajan, Jayaraman J.; Ramamurthy, Karthikeyan Natesan; Spanias, Andreas; Nasiopoulos, Panos.
Proceedings of the 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012. 2012. p. 494-497 6274289.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Learning multilevel dictionaries for compressed sensing using discriminative clustering
AU - Thiagarajan, Jayaraman J.
AU - Ramamurthy, Karthikeyan Natesan
AU - Spanias, Andreas
AU - Nasiopoulos, Panos
PY - 2012
Y1 - 2012
N2 - The performance of sparse recovery using compressed measurements improves when dictionaries learned from training data are used in place of predefined dictionaries. In this paper, we propose to learn incoherent multilevel dictionaries using discriminative clustering in each level. To this end, we present the discriminative K-lines clustering that iterates between identifying the cluster centers and computing the discriminant directions. A scheme for computing representations using the proposed dictionary is also developed. Simulation results for compressed sensing using standard images demonstrate that incorporating incoherence in the dictionary results in improved recovery performance. Furthermore, we implement the proposed algorithms as part of a sparse representations toolbox for the J-DSP software package.
AB - The performance of sparse recovery using compressed measurements improves when dictionaries learned from training data are used in place of predefined dictionaries. In this paper, we propose to learn incoherent multilevel dictionaries using discriminative clustering in each level. To this end, we present the discriminative K-lines clustering that iterates between identifying the cluster centers and computing the discriminant directions. A scheme for computing representations using the proposed dictionary is also developed. Simulation results for compressed sensing using standard images demonstrate that incorporating incoherence in the dictionary results in improved recovery performance. Furthermore, we implement the proposed algorithms as part of a sparse representations toolbox for the J-DSP software package.
KW - compressed sensing
KW - Discriminative clustering
KW - incoherent dictionaries
KW - sparse representations
UR - http://www.scopus.com/inward/record.url?scp=84867178186&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84867178186&partnerID=8YFLogxK
U2 - 10.1109/IIH-MSP.2012.125
DO - 10.1109/IIH-MSP.2012.125
M3 - Conference contribution
AN - SCOPUS:84867178186
SN - 9780769547121
SP - 494
EP - 497
BT - Proceedings of the 2012 8th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2012
ER -