nips nips2001 nips2001-193 nips2001-193-reference knowledge-graph by maker-knowledge-mining
Source: pdf
Author: Yang Song, Luis Goncalves, Pietro Perona
Abstract: This paper presents an unsupervised learning algorithm that can derive the probabilistic dependence structure of parts of an object (a moving human body in our examples) automatically from unlabeled data. The distinguished part of this work is that it is based on unlabeled data, i.e., the training features include both useful foreground parts and background clutter and the correspondence between the parts and detected features are unknown. We use decomposable triangulated graphs to depict the probabilistic independence of parts, but the unsupervised technique is not limited to this type of graph. In the new approach, labeling of the data (part assignments) is taken as hidden variables and the EM algorithm is applied. A greedy algorithm is developed to select parts and to search for the optimal structure based on the differential entropy of these variables. The success of our algorithm is demonstrated by applying it to generate models of human motion automatically from unlabeled real image sequences.
[1] Y. Amit and A. Kong, “Graphical templates for model registration”, IEEE Transactions on Pattern Analysis and Machine Intelligence, 18:225–236, 1996.
[2] C.K. Chow and C.N. Liu, “Approximating discrete probability distributions with dependence trees”, IEEE Transactions on Information Theory, 14:462–467, 1968.
[3] T.M. Cover and J.A. Thomas, Elements of Information Theory, John Wiley and Sons, 1991.
[4] N. Friedman and M. Goldszmidt, “Learning bayesian networks from data”, Technical report, AAAI 1998 Tutorial, http://robotics.stanford.edu/people/nir/tutorial/, 1998.
[5] M.I. Jordan, editor, Learning in Graphical Models, MIT Press, 1999.
[6] M. Meila and M.I. Jordan, “Learning with mixtures of trees”, Journal of Machine Learning Rearch, 1:1–48, 2000.
[7] Y. Song, X. Feng, and P. Perona, “Towards detection of human motion”, In Proc. IEEE CVPR 2000, volume 1, pages 810–817, June 2000.
[8] Y. Song, L. Goncalves, E. Di Bernardo, and P. Perona, “Monocular perception of biological motion in johansson displays”, Computer Vision and Image Understanding, 81:303–327, 2001.
[9] Nathan Srebro, “Maximum likelihood bounded tree-width markov networks”, In UAI, pages 504–511, San Francisco, CA, 2001.
[10] C. Tomasi and T. Kanade, “Detection and tracking of point features”, Tech. Rep. CMU-CS-91132,Carnegie Mellon University, 1991.
[11] M. Weber, M. Welling, and P. Perona, “Unsupervised learning of models for recognition”, In Proc. ECCV, volume 1, pages 18–32, June/July 2000.
[12] Markus Weber, Unsupervised Learning of Models for Object Recognition, Ph.d. thesis, Caltech, May 2000.