Invariant feature set generation with the linear manifold self-organizing map

Authors:
Huicheng Zheng
Affiliations:
School of Information Science and Technology, Sun Yat-sen University, Guangzhou, China
Venue:
ACCV'10 Proceedings of the 10th Asian conference on Computer vision - Volume Part IV
Year:
2010

Citing 8
Cited 0

Locally Weighted Learning

Artificial Intelligence Review - Special issue on lazy learning
Self-organized formation of various invariant-feature filters in the adaptive-subspace SOM

Neural Computation
Self-Organizing Maps

Self-Organizing Maps
Subspace Dimension Selection and Averaged Learning Subspace Method in Handwritten Digit Classification

ICANN 96 Proceedings of the 1996 International Conference on Artificial Neural Networks
Candid Covariance-Free Incremental Principal Component Analysis

IEEE Transactions on Pattern Analysis and Machine Intelligence
Linear manifold topographic map formation based on an energy function with on-line adaptation rules

Neurocomputing
Learning nonlinear manifolds based on mixtures of localized linear manifolds under a self-organizing framework

Neurocomputing
Handwritten digit recognition by adaptive-subspace self-organizing map (ASSOM)

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

One of the most important challenges faced by computer vision is the almost unlimited possibilities of variation associated with the objects. It has been hypothesized that the brain represents image manifolds as manifolds of stable neural-activity patterns. In this paper, we explore the possibility of manifold representation with a set of topographically organized neurons with each representing a local linear manifold and capturing some local linear feature invariance. In particular, we propose to consider the local subspace learning at each neuron of the network from a Gaussian likelihood point of view. Robustness of the algorithm with respect to the learning rate issue is obtained by considering statistical efficiency. Compared to its predecessors, the proposed network is more adaptive and robust in learning globally nonlinear data manifolds, which is verified by experiments on handwritten digit image modeling.