Time Organized Maps Learning cortical topography from spatiotemporal stimuli

1
Time Organized Maps Learning cortical
topography from spatiotemporal stimuli

• Learning cortical topography from spatiotemporal
  stimuli, J. Wiemer, F. Spengler, F. Joublin, P.
  Stagge, S. Wacquant, Biological Cybernetics, 2000
• The Time-Organized Map Algorithm Extending the
  Self-Organizing Map to Spatiotemporal Signals,
  Jan C.Wiemer, Neural Computation, 2003

Presented by Mojtaba Solgi
TexPoint fonts used in EMF. Read the TexPoint
manual before you delete this box. AAAAA
2
Outline

• The purpose and biological motivation
• The Model TOM Algorithm
• Wave propagation
• Learning
• Experiments and Results
• Gaussian stimuli
• Generic artificial stimuli
• Semi-natural stimuli
• Discussion
• z

3
Neurobiological experiments, Spengler et al.,
1996, 1999
4
Terminology

• Integration
• Fusion of different stimuli into one
  representation
• Segregation
• Process of Increasing representational distance
  of different stimuli
• z

5
2D Network Architecture Activation positional
shift
6
One-dimensional model
7
Wave propagation
8
Integration and Segregation
9
Algorithm

• Compute neurons activations and the position of
  the top winner neuron
• Compute the neural position of the propagated
  wave from the last time step activation

10
Algorithm Cont.

• Shift the position of the top winner neuron due
  to interaction with propagated wave

11
Algorithm Cont.

• Again shift the position of the winner neuron
  this time due to noise
• Update the winner neurons weights
• SOM
• Hebbian

12
Experiments with Gaussian stimuli 2D neural
layer

• Simulation of ontogenesis (Development)

13
Experiments with Gaussian stimuli 2D neural
layer

• Simulation of post-ontogenetic plasticity

14
One-dimensional model
15
Experiments with generic artificial stimuli 1D
neural layer
The input
16
Experiments with semi-natural stimuli 1D neural
layer
17
Experiments with semi-natural stimuli 1D neural
layer
18
Discussion

• Importance of temporal stimulus for development
  of cortical topography
• Continuous mapping of related stimuli
• Inter-Stimulus-Interval-Dependant representations
• Hardly scalable
• No recognition performance on real-world problems
• Tested only on artificial input

19
Summary

• Utilizing temporal information in developing
  cortical topography
• Wave-like spread of cortical activity
• Experiments and results show compatibility of the
  model with neurobiological observations
• Biologically inspired and plausible, but no
  engineering performance

20
Thank you!

• Any thoughts/question?