Labeled images verification using Gaussian mixture models

We are proposing in this paper an automated system to verify that images are correctly associated to labels. The novelty of the system is in the use of Gaussian Mixture Models (GMMs) as statistical modeling scheme as well as in several improvements introduced specifically for the verification task. Our approach is evaluated using the Caltech 101 database. Starting from an initial baseline system providing an equal error rate of 27.4%, we show that the rate of errors can be reduced down to 13% by introducing several optimizations of the system. The advantage of the approach lies in the fact that basically any object can be generically and blindly modeled with limited supervision. A potential target application could be a post-filtering of images returned by search engines to prune out or reorder less relevant images.

[1]  Nello Cristianini,et al.  An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[2]  Rolf Ingold,et al.  Modelling Spoken Signatures with Gaussian Mixture Model Adaptation , 2007, 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07.

[3]  Kuldip K. Paliwal,et al.  Polynomial features for robust face authentication , 2002, Proceedings. International Conference on Image Processing.

[4]  Kuldip K. Paliwal,et al.  Fast features for face authentication under illumination direction changes , 2003, Pattern Recognit. Lett..

[5]  R. Lippmann.  Pattern classification using neural networks , 1989, IEEE Communications Magazine.

[6]  Hermann Ney,et al.  FIRE - Flexible Image Retrieval Engine: ImageCLEF 2004 Evaluation , 2004, CLEF.

[7]  Luc Van Gool,et al.  Efficient Mining of Frequent and Distinctive Feature Configurations , 2007, 2007 IEEE 11th International Conference on Computer Vision.

[8]  Hermann Ney,et al.  Automatic categorization of medical images for content-based retrieval and data mining. , 2005, Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society.

[9]  Thomas S. Huang,et al.  Image processing , 1971 .

[10]  Pietro Perona,et al.  Learning Generative Visual Models from Few Training Examples: An Incremental Bayesian Approach Tested on 101 Object Categories , 2004, 2004 Conference on Computer Vision and Pattern Recognition Workshop.

[11]  Jitendra Malik,et al.  SVM-KNN: Discriminative Nearest Neighbor Classification for Visual Category Recognition , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[12]  Douglas A. Reynolds,et al.  Speaker Verification Using Adapted Gaussian Mixture Models , 2000, Digit. Signal Process..

[13]  Andrew B. Watson,et al.  Image Compression Using the Discrete Cosine Transform , 1994 .

[14]  T.R. Lim,et al.  Car recognition using Gabor filter feature extraction , 2002, Asia-Pacific Conference on Circuits and Systems.

[15]  T. Moon.  The expectation-maximization algorithm , 1996, IEEE Signal Process. Mag..

[16]  Samy Bengio,et al.  User authentication via adapted statistical models of face images , 2006, IEEE Transactions on Signal Processing.

[17]  Belur V. Dasarathy,et al.  Nearest neighbor (NN) norms: NN pattern classification techniques , 1991 .