## S-114.204 Modelling of Perception | |

- 1. Pattern theory: a unifying perspective (D. Mumford)
- mathematical formulation of the visual perception porblem (MAP/MDL)
- basic signal transforms (noise,superposition,warp,interruption)
- iterative (top-down) solution

- 2. Modal structure and reliable inference (A. Jepson, W. Richards)
- use of regular (known) structures in interpreting the perception
- eg. preditable bouncing of a ball in a box

- 3. Priors, preferences and categorical percepts, L
(W. Richards, A. Jepson, J. Feldman)
- selection of most regular or least special solution that fits the observations
- based on discrete set of models {Si} matching the image I

- 4. Bayesian decision theory and psychophysics, L (A. L. Yuille, H.H. Bülthoff)
- integration of several visual cues
- psykophysical evidence of the integration process

- 5. Observer theory, bayes theory and psychophysics, L
(B.M. Bennett, D. D. Hoffman, C. Prakash, S.N. Richman)
- generalized Bayes rule (posterior probability) for cases when p(I)~0
- Note: the formulation of the problem is based on misconception of the normalizing term in the Bayes rule. The reasoning and the results may still be useful.

- 6. Implications of a Bayesian formulation of visual information
processing for psychophysics (D. C. Knill, D. Kersten, P. Mamassian)
- Bayesian perception model for inferring the form of a surface

- 7. Shape from texture: ideal observers and human psychophysics,L (A. Blake, H.H. Bülthoff, D. Sheinberg)
- 8. A computational theory for binocular stereopsis, L (P. N. Belhumeur)
- Bayesian theory for stereo vision

- 9. A generic viewpoint assumption in bayesian framework (W. T. Freeman)
- shape from shading technique
- measure of goodness (a priori probability) is the robustness of the solution to changes in viewing angle

- 10. Experiencing and perceiving visual surfaces (K. Nakayama, S. Shimojo)
- ecological optics approach
- similar to previous, generic viewpoint, principle

- 11. The perception of shading and reflectance (E. H. Adelson, A. P. Pentland)
- prior probability defined by complexity of constructing the surface
- included elements are form of the surface, lighting and color

- 12. Banishing the homunculus (H. Barlow)
- sketch for computational goal of the function of the cortex
- goal is to remove the homunculus from the theories (that is, a small human in the brain that performs the final perception task from sensory maps etc.
- See also Barlow's recent review and references therein.

- Edward H. Adelson
- Peter N. Belhumeur
- Andrew Blake
- Heinrich H. Bülthoff
- Walter J Freeman
- Donald D. Hoffman
- Allan Jepson
- Daniel Kersten
- David C. Knill
- David Mumford
- Ken Nakayama
- Whitman Richards
- Alan L. Yuille

- Information processing in visual cortex (V1, Simple/Complex cells)
- Sparse coding with an overcomplete basis set: a strategy employed by V1? (1996) Olshausen BA, Field DJ. S Vision Research PostScript
- Emergence of Phase and Shift Invariant Features by Decomposition of Natural Images into Independent Feature Subspaces, A. Hyvärinen and P. O. Hoyer, Neural Computation 12(7):1705-1720, 2000. PostScript