This Article 
   
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
Data Driven Image Models through Continuous Joint Alignment
February 2006 (vol. 28 no. 2)
pp. 236-250
ASCII Text
x 
 
Erik G. Learned-Miller, "Data Driven Image Models through Continuous Joint Alignment," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 28, no. 2, pp. 236-250, February, 2006.
 
 
BibTex
x
 
@article{ 10.1109/TPAMI.2006.34,
author = {Erik G. Learned-Miller},
title = {Data Driven Image Models through Continuous Joint Alignment},
journal ={IEEE Transactions on Pattern Analysis and Machine Intelligence},
volume = {28},
number = {2},
issn = {0162-8828},
year = {2006},
pages = {236-250},
doi = {http://doi.ieeecomputersociety.org/10.1109/TPAMI.2006.34},
publisher = {IEEE Computer Society},
address = {Los Alamitos, CA, USA},
}
 
 
RefWorks Procite/RefMan/Endnote
x 
 
TY - JOUR
JO - IEEE Transactions on Pattern Analysis and Machine Intelligence
TI - Data Driven Image Models through Continuous Joint Alignment
IS - 2
SN - 0162-8828
SP236
EP250
EPD - 236-250
A1 - Erik G. Learned-Miller,
PY - 2006
KW - Index Terms- Alignment
KW - artifact removal
KW - bias removal
KW - congealing
KW - clustering
KW - correspondence
KW - density estimation
KW - entropy
KW - maximum likelihood
KW - medical imaging
KW - magnetic resonance imaging
KW - nonparametric statistics
KW - registration
KW - unsupervised learning.
VL - 28
JA - IEEE Transactions on Pattern Analysis and Machine Intelligence
ER -
 
This paper presents a family of techniques that we call congealing for modeling image classes from data. The idea is to start with a set of images and make them appear as similar as possible by removing variability along the known axes of variation. This technique can be used to eliminate "nuisance” variables such as affine deformations from handwritten digits or unwanted bias fields from magnetic resonance images. In addition to separating and modeling the latent images—i.e., the images without the nuisance variables—we can model the nuisance variables themselves, leading to factorized generative image models. When nuisance variable distributions are shared between classes, one can share the knowledge learned in one task with another task, leading to efficient learning. We demonstrate this process by building a handwritten digit classifier from just a single example of each class. In addition to applications in handwritten character recognition, we describe in detail the application of bias removal from magnetic resonance images. Unlike previous methods, we use a separate, nonparametric model for the intensity values at each pixel. This allows us to leverage the data from the MR images of different patients to remove bias from each other. Only very weak assumptions are made about the distributions of intensity values in the images. In addition to the digit and MR applications, we discuss a number of other uses of congealing and describe experiments about the robustness and consistency of the method.

[1] B.J. Frey and N. Jojic, “Transformation-Invariant Clustering and Dimensionality Reduction Using EM,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 25, no. 1, pp. 1-17, Jan. 2003.
[2] E. Miller, N. Matsakis, and P. Viola, “Learning from One Example through Shared Densities on Transforms,” Proc. IEEE Computer Vision and Pattern Recognition Conf., 2000.
[3] E.G. Miller, “Learning from One Example in Machine Vision by Sharing Probability Densities,” PhD thesis, Massachusetts Inst. of Tech nology, 2002, http://www.cs.umass.edu/~elm/papersthesis.pdf .
[4] T.M. Cover and J.A. Thomas, Elements of Information Theory. John Wiley & Sons, 1991.
[5] E.G. Miller and C. Chefd'hotel, “Practical Non-Parametric Density Estimation on a Transformation Group for Vision,” Proc. IEEE Computer Vision and Pattern Recognition Conf., 2003.
[6] D.P. Huttenlocher, G.A. Klanderman, and W.J. Rucklidge, “Comparing Images Using the Hausdorff Distance,” Trans. IEEE Pattern Analysis and Machine Intelligence, vol. 15, no. 9, pp. 850-863, Sept. 1993.
[7] A.C. Fan, “A Variational Approach to MR Bias Correction,” MS thesis, Massachusetts Inst. of Tech nology, 2003.
[8] W.M. Wells, W.E.L. Grimson, R. Kikinis, and F. Jolesz, “Adaptive Segmentation of MRI Data,” IEEE Trans. Medical Imaging, vol. 15, pp. 429-442, 1996.
[9] P.A. Viola, “Alignment by Maximization of Mutual Information,” PhD thesis, Massachusetts Inst. of Tech nology, 1995, ftp://publications.ai.mit. edu/ai-publications/ pdfAITR-1548.pdf.
[10] P. Viola and W.M. Wells III, “Mutual Information: An Approach for the Registration of Object Models and Images,” Int'l J. Computer Vision, 1997.
[11] O. Vasicek, “A Test for Normality Based on Sample Entropy,” J. Royal Statistical Soc., Series B, vol. 38, no. 1, pp. 54-59, 1976.
[12] E.G. Learned-Miller and J.W. Fisher, “ICA Using Spacings Estimates of Entropy,” J. Machine Learning Research, vol. 4, pp. 1271-1295, 2003.
[13] D.L. Collins, A.P. Zijdenbos, J.G. Kollokian, N.J. Sled, C.J. Kabani, C.J. Holmes, and A.C. Evans, “Design and Construction of a Realistic Digital Brain Phantom,” IEEE Trans. Medical Imaging, vol. 17, pp. 463-468, 1998.
[14] M. Revow, C. Williams, and G. Hinton, “Using Generative Models for Handwritten Digit Recognition,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 18, no. 6, pp. 592-606, June 1996.
[15] T.F. Cootes, G.J. Edwards, and C.J. Taylor, “Active Appearance Models,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 23, no. 6, pp. 681-685, June 2001.
[16] M. Jones and T. Poggio, “Model-Based Matching by Linear Combinations of Prototypes,” Technical Report AI Memo 1583, Massachusetts Inst. of Tech nology, 1995, ftp://publications.ai. mit.edu/ai-publications/ pdfAIM-1583.pdf.
[17] T. Vetter, M. Jones, and T. Poggio, “A Bootstrapping Algorithm for Learning Linear Models of Object Classes,” Proc. IEEE Computer Vision and Pattern Recognition Conf., pp. 40-46, 1997.
[18] B. Frey and N. Jojic, “Estimating Mixture Models of Images and Inferring Spatial Transformations Using the EM Algorithm,” Proc. IEEE Computer Vision and Pattern Recognition Conf., pp. 416-422, 1999.
[19] B. Frey and N. Jojic, “Transformed Component Analysis: Joint Estimation of Spatial Transformations and Image Components,” Proc. Int'l Conf. Computer Vision, 1999.
[20] C. Bishop, M. Svensén, and C. Williams, “GTM: The Generative Topographic Mapping,” Neural Computation, vol. 10, no. 1, pp. 215-234, 1998.
[21] P. Simard, Y. LeCun, and J. Denker, “Efficient Pattern Recognition Using a New Transformation Distance,” Proc. Advanced in Neural Information Processing Systems 5, pp. 51-58, 1993.
[22] N. Vasconcelos and A. Lippman, “Multiresolution Tangent Distance for Affine-Invariant Classification,” Proc. Advanced in Neural Information Processing Systems 10, 1998.

Index Terms:
Index Terms- Alignment, artifact removal, bias removal, congealing, clustering, correspondence, density estimation, entropy, maximum likelihood, medical imaging, magnetic resonance imaging, nonparametric statistics, registration, unsupervised learning.
Citation:
Erik G. Learned-Miller, "Data Driven Image Models through Continuous Joint Alignment," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 28, no. 2, pp. 236-250, Feb. 2006, doi:10.1109/TPAMI.2006.34
Usage of this product signifies your acceptance of the Terms of Use.