论文信息 - Activity bubbles and natural image sequences

Activity bubbles and natural image sequences

Abstract. Recently, different models of the statistical structure of natural images have been proposed. Maximizing sparseness, or alternatively temporal coherence of linear filter outputs leads to the emergence of simple cell properties. Taking account of the basic dependencies of linear filter outputs enables modelling of complex cell and topographic properties as well. In this paper, I propose a unifying framework for all these statistical properties, based on the concept of spatio-temporal activity bubbles.

A. Hyvärinen | Aapo Hyvärinen

[1] Aapo Hyvärinen,et al. Receptive Fields Similar to Simple Cells Maximize Temporal Coherence in Natural Video , 2002, ICANN.

[2] Terrence J. Sejnowski,et al. Slow Feature Analysis: Unsupervised Learning of Invariances , 2002, Neural Computation.

[3] Aapo Hyvärinen,et al. A Novel Temporal Generative Model of Natural Video as an Internal Model in Early Vision , 2002 .

[4] Aapo Hyvärinen,et al. A two-layer sparse coding model learns simple and complex cell receptive fields and topography from natural images , 2001, Vision Research.

[5] Aapo Hyvärinen,et al. Topographic Independent Component Analysis , 2001, Neural Computation.

[6] David J. Field,et al. Emergence of simple-cell receptive field properties by learning a sparse code for natural images , 1996, Nature.

[7] David J. Field,et al. What Is the Goal of Sensory Coding? , 1994, Neural Computation.

[8] Peter Földiák,et al. Learning Invariance from Transformation Sequences , 1991, Neural Comput..