Constructing visual models with a latent space approach

Authors:
Florent Monay;Pedro Quelhas;Daniel Gatica-Perez;Jean-Marc Odobez
Affiliations:
IDIAP Research Institute, Martigny, Switzerland;IDIAP Research Institute, Martigny, Switzerland;IDIAP Research Institute, Martigny, Switzerland;IDIAP Research Institute, Martigny, Switzerland
Venue:
SLSFS'05 Proceedings of the 2005 international conference on Subspace, Latent Structure and Feature Selection
Year:
2005

Citing 12
Cited 1

Unsupervised learning by probabilistic latent semantic analysis

Machine Learning
Modern Information Retrieval

Modern Information Retrieval
Object Recognition as Machine Translation: Learning a Lexicon for a Fixed Image Vocabulary

ECCV '02 Proceedings of the 7th European Conference on Computer Vision-Part IV
Variational Extensions to EM and Multinomial PCA

ECML '02 Proceedings of the 13th European Conference on Machine Learning
Content-Based Image Retrieval Based on Local Affinely Invariant Regions

VISUAL '99 Proceedings of the Third International Conference on Visual Information and Information Systems
Latent dirichlet allocation

The Journal of Machine Learning Research
Video Google: A Text Retrieval Approach to Object Matching in Videos

ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
On image auto-annotation with latent space models

MULTIMEDIA '03 Proceedings of the eleventh ACM international conference on Multimedia
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision
PLSA-based image auto-annotation: constraining the latent space

Proceedings of the 12th annual ACM international conference on Multimedia
Modeling Scenes with Local Descriptors and Latent Aspects

ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1 - Volume 01
Image classification for content-based indexing

IEEE Transactions on Image Processing

Contextual classification of image patches with latent aspect models

Journal on Image and Video Processing - Special issue on patches in vision

Quantified Score

Hi-index	0.00

Visualization

Abstract

We propose the use of latent space models applied to local invariant features for object classification. We investigate whether using latent space models enables to learn patterns of visual co-occurrence and if the learned visual models improve performance when less labeled data are available. We present and discuss results that support these hypotheses. Probabilistic Latent Semantic Analysis (PLSA) automatically identifies aspects from the data with semantic meaning, producing unsupervised soft clustering. The resulting compact representation retains sufficient discriminative information for accurate object classification, and improves the classification accuracy through the use of unlabeled data when less labeled training data are available. We perform experiments on a 7-class object database containing 1776 images.