Details
| Original language | English |
|---|---|
| Title of host publication | Computer Analysis of Images and Patterns - 14th International Conference, CAIP 2011, Proceedings |
| Pages | 410-419 |
| Number of pages | 10 |
| Edition | PART 1 |
| Publication status | Published - 2011 |
| Event | 14th International Conference on Computer Analysis of Images and Patterns, CAIP 2011 - Seville, Spain Duration: 29 Aug 2011 → 31 Aug 2011 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Number | PART 1 |
| Volume | 6854 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (electronic) | 1611-3349 |
Abstract
In this paper we propose to enhance the training data of boosting-based object detection frameworks by the use of principal component analysis (PCA). The quality of boosted classifiers highly depends on the image databases exploited in training. We observed that negative training images projected into the objects PCA space are often far away from the object class. This broad boundary between the object classes in training can yield to a high classification error of the boosted classifier in the testing phase. We show that transforming the negative training database close to the positive object class can increase the detection performance. In experiments on face detection and the analysis of microscopic cell images, our method decreases the amount of false positives while maintaining a high detection rate. We implemented our approach in a Viola & Jones object detection framework using AdaBoost to combine Haar-like features. But as a preprocessing step our method can easily be integrated in all boosting-based frameworks without additional overhead.
ASJC Scopus subject areas
- Mathematics(all)
- Theoretical Computer Science
- Computer Science(all)
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Computer Analysis of Images and Patterns - 14th International Conference, CAIP 2011, Proceedings. PART 1. ed. 2011. p. 410-419 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6854 LNCS, No. PART 1).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - PCA enhanced training data for adaboost
AU - Ehlers, Arne
AU - Baumann, Florian
AU - Spindler, Ralf
AU - Glasmacher, Birgit
AU - Rosenhahn, Bodo
N1 - Funding information: This work has been partially funded by the DFG within the excellence cluster REBIRTH.
PY - 2011
Y1 - 2011
N2 - In this paper we propose to enhance the training data of boosting-based object detection frameworks by the use of principal component analysis (PCA). The quality of boosted classifiers highly depends on the image databases exploited in training. We observed that negative training images projected into the objects PCA space are often far away from the object class. This broad boundary between the object classes in training can yield to a high classification error of the boosted classifier in the testing phase. We show that transforming the negative training database close to the positive object class can increase the detection performance. In experiments on face detection and the analysis of microscopic cell images, our method decreases the amount of false positives while maintaining a high detection rate. We implemented our approach in a Viola & Jones object detection framework using AdaBoost to combine Haar-like features. But as a preprocessing step our method can easily be integrated in all boosting-based frameworks without additional overhead.
AB - In this paper we propose to enhance the training data of boosting-based object detection frameworks by the use of principal component analysis (PCA). The quality of boosted classifiers highly depends on the image databases exploited in training. We observed that negative training images projected into the objects PCA space are often far away from the object class. This broad boundary between the object classes in training can yield to a high classification error of the boosted classifier in the testing phase. We show that transforming the negative training database close to the positive object class can increase the detection performance. In experiments on face detection and the analysis of microscopic cell images, our method decreases the amount of false positives while maintaining a high detection rate. We implemented our approach in a Viola & Jones object detection framework using AdaBoost to combine Haar-like features. But as a preprocessing step our method can easily be integrated in all boosting-based frameworks without additional overhead.
UR - http://www.scopus.com/inward/record.url?scp=80052822150&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-23672-3_50
DO - 10.1007/978-3-642-23672-3_50
M3 - Conference contribution
AN - SCOPUS:80052822150
SN - 9783642236716
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 410
EP - 419
BT - Computer Analysis of Images and Patterns - 14th International Conference, CAIP 2011, Proceedings
T2 - 14th International Conference on Computer Analysis of Images and Patterns, CAIP 2011
Y2 - 29 August 2011 through 31 August 2011
ER -