learning multi sensory integration with self organization
play

Learning Multi-Sensory Integration with Self- Organization and - PowerPoint PPT Presentation

Learning Multi-Sensory Integration with Self- Organization and Statistics Johannes Bauer, Stefan Wermter http://www.informatik.uni-hamburg.de/WTM/ The Superior Colliculus Johannes Bauer Multi-Sensory Integration through Self-Organization and


  1. Learning Multi-Sensory Integration with Self- Organization and Statistics Johannes Bauer, Stefan Wermter http://www.informatik.uni-hamburg.de/WTM/

  2. The Superior Colliculus Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 2

  3. The Superior Colliculus inverse effectiveness [1] spatial principle [1] sub-additive coincident neural response close additive disparate super-additive uni-sensory baseline optimal integration [2] Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 3

  4. What’s Interesting About That? ? Place in SC Meaning of Input Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 4

  5. The Algorithm [3] 𝜍: input stimulus 𝑏 𝑙 : activity of 𝑗 𝑙 𝜍 π‘š : preferred value of 𝑝 π‘š Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 5

  6. The Algorithm [3] 𝜍: input stimulus 𝑏 𝑙 : activity of 𝑗 𝑙 𝜍 π‘š : preferred value of 𝑝 π‘š ∼ 𝑄 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 𝜍 = 𝜍 π‘š ) 𝑄 𝜍 = 𝜍 π‘š 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 𝑄(𝜍 = 𝜍 π‘š ) 𝑄 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 6

  7. The Algorithm [3] 𝜍: input stimulus 𝑏 𝑙 : activity of 𝑗 𝑙 𝜍 π‘š : preferred value of 𝑝 π‘š ∼ 𝑄 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 𝜍 = 𝜍 π‘š ) 𝑄 𝜍 = 𝜍 π‘š 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 𝑄(𝜍 = 𝜍 π‘š ) 𝑄 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 6

  8. The Algorithm [3] 𝜍: input stimulus 𝑏 𝑙 : activity of 𝑗 𝑙 𝜍 π‘š : preferred value of 𝑝 π‘š ∼ 𝑄(𝑏 𝑙 ∣ 𝜍 = 𝜍 π‘š ) 𝑙 𝑄 𝜍 = 𝜍 π‘š 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 𝑄(𝜍 = 𝜍 π‘š ) 𝑄(𝑏 𝑙 ) 𝑙 Noise independent. Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 6

  9. The Algorithm [3] 𝜍: input stimulus 𝑏 𝑙 : activity of 𝑗 𝑙 𝜍 π‘š : preferred value of 𝑝 π‘š ∼ 𝑄(𝑏 𝑙 ∣ 𝜍 = 𝜍 π‘š ) 𝑙 𝑄 𝜍 = 𝜍 π‘š 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 𝑄(𝑏 𝑙 ) 𝑙 Noise independent. 𝜍 unif. dist. Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 6

  10. The Algorithm [3] 𝜍: input stimulus 𝑏 𝑙 : activity of 𝑗 𝑙 𝜍 π‘š : preferred value of 𝑝 π‘š 𝑄 𝜍 = 𝜍 π‘š 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 ∼ 𝑄(𝑏 𝑙 ∣ 𝜍 = 𝜍 π‘š ) 𝑙 Noise independent. 𝜍 unif. dist. Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 6

  11. The Algorithm [3] 𝑄 𝜍 = 𝜍 π‘š 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 ∼ 𝑄(𝑏 𝑙 ∣ 𝜍 = 𝜍 π‘š ) 𝑙 Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 11

  12. The Algorithm [3] can be adapted SOM-like π‘π‘œ π‘˜,𝑏 [𝑏 𝑙 ] Assume 𝑄 𝑏 𝑙 ∣ 𝜍 = 𝜍 π‘˜ = βˆ‘π‘π‘œ π‘˜,𝑏 , then 𝑄 𝜍 = 𝜍 π‘˜ ∣ 𝑏 1 , 𝑏 2 , … , 𝑏 𝑛 ∼ π‘π‘œ π‘˜,π‘š 𝑏 π‘š βˆ‘ π‘π‘œ π‘˜,π‘š π‘š Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 12

  13. Comparison to regular SOM Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 13

  14. Comparison to regular SOM Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 13

  15. The Network in Action β€˜visual’ input β€˜auditory’ input population-coded PDF Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 15

  16. The Network in Action β€˜visual’ input β€˜auditory’ input population-coded PDF Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 16

  17. Performance The network integrates multi-sensory information. errors given β€˜visual’ input errors given β€˜auditory’ input errors given multi-sensory input The network replicates biological phenomena. Spatial Principle Inverse Effectiveness Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 17

  18. Conclusion presented a novel self-organizing ANN algorithm which  learns to combine information near-optimally  shows spatial principle and MLE-like behavior  shows benefit of multisensory integration  learns to compute a PDF for latent variables  is unsupervised  has few inbuilt assumptions Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 18

  19. The End References: [1]: Stanford, T. R., Quessy, S., Stein, B. E., Jul. 2005. Evaluating the operations underlying multisensory integration in the cat superior colliculus. The Journal of Neuroscience 25 (28), 6499 – 6508. [2]: Alais, D., Burr, D., Feb. 2004. The ventriloquist effect results from Near-Optimal bimodal integration . Current Biology 14 (3), 257 – 262. [3]: Bauer, J. and Wermter, S., Sept. 2013. Self-organized neural learning of statistical inference from high-dimensional data. In: Proceedings of the International Joint Conference on Artificial Intelligence 2013 . Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 14

  20. Performance – Behavioral * *simulation parameters differ from rest of talk. auditory 𝜏 𝑏 = 4.594 βˆ™ 10 βˆ’4 π‘ž 𝑏 = 1.680 βˆ™ 10 βˆ’1 visual 𝜏 𝑀 = 1.061 βˆ™ 10 βˆ’4 π‘ž 𝑀 = 8.272 βˆ™ 10 βˆ’1 optimal 1 β‰ˆ 4.185 βˆ™ 10 βˆ’5 𝜏 𝑛,π‘π‘žπ‘’ = 1 2 + 1 2 πœπ‘€ πœπ‘ multi-sensory π‘ž 𝑏,π‘π‘žπ‘’ β‰ˆ 1.876 βˆ™ 10 βˆ’1 𝜏 𝑛 = 8.153 βˆ™ 10 βˆ’5 π‘ž 𝑀,π‘π‘žπ‘’ β‰ˆ 8.124 βˆ™ 10 βˆ’1 Johannes Bauer Multi-Sensory Integration through Self-Organization and Statistics 20

Recommend


More recommend