论文信息 - Extracting Slow Subspaces from Natural Videos Leads to Complex Cells

Extracting Slow Subspaces from Natural Videos Leads to Complex Cells

Natural videos obtained from a camera mounted on a cat's head are used as stimuli for a network of subspace energy detectors. The network is trained by gradient ascent on an objective function defined by the squared temporal derivatives of the cells' outputs. The resulting receptive fields are invariant to both contrast polarity and translation and thus resemble complex type receptive fields.

[1] S. Grossberg. A neural model of attention, reinforcement and discrimination learning. , 1975, International review of neurobiology.

[2] E. Oja. Simplified neuron model as a principal component analyzer , 1982, Journal of mathematical biology.

[3] Kunihiko Fukushima,et al. Neocognitron: A hierarchical neural network capable of visual pattern recognition , 1988, Neural Networks.

[4] Geoffrey E. Hinton. Connectionist Learning Procedures , 1989, Artif. Intell..

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

[6] Teuvo Kohonen,et al. Emergence of invariant-feature detectors in the adaptive-subspace self-organizing map , 1996, Biological Cybernetics.

[7] T. Poggio,et al. Hierarchical models of object recognition in cortex September 23 , 1999 , 1999 .

[8] Konrad Paul Kording,et al. Learning with two sites of synaptic integration , 2000, Network.

[9] Aapo Hyvärinen,et al. Emergence of Phase- and Shift-Invariant Features by Decomposition of Natural Images into Independent Feature Subspaces , 2000, Neural Computation.