Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Pattern Recognition |
| Untertitel | 33rd DAGM Symposium, Proceedings |
| Seiten | 81-90 |
| Seitenumfang | 10 |
| Publikationsstatus | Veröffentlicht - 2011 |
| Veranstaltung | 33rd Annual Symposium of German Pattern Recognition Association, DAGM 2011 - Frankfurt/Main, Deutschland Dauer: 31 Aug. 2011 → 2 Sept. 2011 |
Publikationsreihe
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Band | 6835 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (elektronisch) | 1611-3349 |
Abstract
Many challenging computer vision problems can be formulated as a multilinear model. Classical methods like principal component analysis use singular value decomposition to infer model parameters. Although it can solve a given problem easily if all measurements are known this prerequisite is usually violated for computer vision applications. In the current work, a standard tool to estimate singular vectors under incomplete data is reformulated as an energy minimization problem. This admits for a simple and fast gradient descent optimization with guaranteed convergence. Furthermore, the energy function is generalized by introducing an L 2-regularization on the parameter space. We show a quantitative and qualitative evaluation of the proposed approach on an application from structure-from-motion using synthetic and real image data, and compare it with other works.
ASJC Scopus Sachgebiete
- Mathematik (insg.)
- Theoretische Informatik
- Informatik (insg.)
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Pattern Recognition: 33rd DAGM Symposium, Proceedings. 2011. S. 81-90 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 6835 LNCS).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Multilinear model estimation with L2-regularization
AU - Schmidt, Frank R.
AU - Ackermann, Hanno
AU - Rosenhahn, Bodo
PY - 2011
Y1 - 2011
N2 - Many challenging computer vision problems can be formulated as a multilinear model. Classical methods like principal component analysis use singular value decomposition to infer model parameters. Although it can solve a given problem easily if all measurements are known this prerequisite is usually violated for computer vision applications. In the current work, a standard tool to estimate singular vectors under incomplete data is reformulated as an energy minimization problem. This admits for a simple and fast gradient descent optimization with guaranteed convergence. Furthermore, the energy function is generalized by introducing an L 2-regularization on the parameter space. We show a quantitative and qualitative evaluation of the proposed approach on an application from structure-from-motion using synthetic and real image data, and compare it with other works.
AB - Many challenging computer vision problems can be formulated as a multilinear model. Classical methods like principal component analysis use singular value decomposition to infer model parameters. Although it can solve a given problem easily if all measurements are known this prerequisite is usually violated for computer vision applications. In the current work, a standard tool to estimate singular vectors under incomplete data is reformulated as an energy minimization problem. This admits for a simple and fast gradient descent optimization with guaranteed convergence. Furthermore, the energy function is generalized by introducing an L 2-regularization on the parameter space. We show a quantitative and qualitative evaluation of the proposed approach on an application from structure-from-motion using synthetic and real image data, and compare it with other works.
UR - http://www.scopus.com/inward/record.url?scp=80053024337&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-23123-0_9
DO - 10.1007/978-3-642-23123-0_9
M3 - Conference contribution
AN - SCOPUS:80053024337
SN - 9783642231223
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 81
EP - 90
BT - Pattern Recognition
T2 - 33rd Annual Symposium of German Pattern Recognition Association, DAGM 2011
Y2 - 31 August 2011 through 2 September 2011
ER -