Abstract
A model for unsupervised learning of visual invariances is presented. The learning method involves a linear combination of anti-Hebbian and Hebbian weight changes, over short and long time scales, respectively. The model is demonstrated on the problem of estimating sub-pixel stereo disparity from a temporal sequence of unprocessed image pairs. After learning on a given image sequence, the model’s ability to detect sub-pixel disparity generalises, without additional learning, to image pairs from other sequences.
The author is a joint member of the Schools of Biological Sciences, and Cognitive and Computing Sciences at the University of Sussex.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
S Becker. Learning to categorize objects using temporal coherence. Neural Information Processing Systems, pages 361–368, 1992.
S Becker and GE Hinton. Self-organizing neural network that discovers surfaces in random-dot stereograms. Nature, 335:161–163, 1992.
SA Clark, A Allard, WM Jenkins, and MM Merzenich. Receptive fields in the body-surface map in adult cortex defined by temporally correlated inputs. Nature, 332:444–445, March 1988.
R Durbin and D Willshaw. An analogue approach to the travelling salesman problem using an elastic net method. Nature, 326(6114):689–691, 1987.
JJ Hopfield. Neurons with graded response have collective computational properties like those of two-state neurons. Proc Nat Ac Sci., 81:3088–3092, 1984.
S Kirkpatrick, CD Gelat, and MP Vecchi. Optimization by simulated annealing. Science, 220(4598):671–680, May 1983.
D Marr. Vision. Freeman, New York, 1982.
GF Poggio. Cortical neural mechanisms of stereopsis studied with dynamic random-dot stereograms. Cold Spring Harbour Symposia on Quantitative Biology, LV:749–756, 1990.
DE Rumelhart, GE Hinton, and RJ Williams. Learning representations by back-propagating errors. Nature, 323:533–536, 1986.
JV Stone. The optimal elastic net: Finding solutions to the travelling salesman problem. ICANN92, pages 170–174, 1992.
JV Stone. Shape from local and global analysis of texture. Phil. Trans. Roy. Soc. Lond.(B), 339(1287):53–65, January 1992.
JV Stone. Computer vision: What is the object? In Prospects for AI, Proc. Artificial Intelligence and Simulation of Behaviour, Birmingham, England. IOS Press, Amsterdam., pages 199–208, April 1993.
G Westheimer. The ferrier lecture, 1992. seeing depth with two eyes: Stereopsis. Proc Royal Soc London, 5, 257:205–214, 1994.
RS Zemel and GE Hinton. Discovering viewpoint invariant relationships that characterize objects. Technical Report, Dept. of Computer Science, University of Toronto, Toronto, ONT MS5 1A4, 1991, 1991.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1995 Springer-Verlag London
About this paper
Cite this paper
Stone, J.V. (1995). Learning Invariances Via Spatio-Temporal Constraints. In: Smith, L.S., Hancock, P.J.B. (eds) Neural Computation and Psychology. Workshops in Computing. Springer, London. https://doi.org/10.1007/978-1-4471-3579-1_6
Download citation
DOI: https://doi.org/10.1007/978-1-4471-3579-1_6
Publisher Name: Springer, London
Print ISBN: 978-3-540-19948-9
Online ISBN: 978-1-4471-3579-1
eBook Packages: Springer Book Archive