Processing Images by Semi-Linear Predictability Minimization

N. N. Schraudolph, M. Eldracher, and J. Schmidhuber. Processing Images by Semi-Linear Predictability Minimization. Network: Computation in Neural Systems, 10(2):133–169, 1999.

Download

1.6MB	1.6MB

Abstract

In the predictability minimization approach, input patterns are fed into a system consisting of adaptive, initially unstructured feature detectors. There are also adaptive predictors constantly trying to predict current feature detector outputs from other feature detector outputs. Simultaneously, however, the feature detectors try to become as unpredictable as possible, resulting in a co-evolution of predictors and feature detectors. This paper describes the implementation of a visual processing system trained by semi-linear predictability minimization, and presents many experiments that examine its response to artificial and real-world images. In particular, we observe that under a wide variety of conditions, predictability minimization results in the development of well-known visual feature detectors.

BibTeX Entry

@article{SchEldSch99,
     author = {Nicol N. Schraudolph and Martin Eldracher
               and J\"urgen Schmid\-huber},
      title = {\href{http://nic.schraudolph.org/pubs/SchEldSch99.pdf}{
               Processing Images by Semi-Linear Predictability Minimization}},
      pages = {133--169},
    journal = {Network: Computation in Neural Systems},
     volume =  10,
     number =  2,
       year =  1999,
   b2h_type = {Journal Papers},
  b2h_topic = {Unsupervised Learning},
   abstract = {
    In the predictability minimization approach, input patterns are
    fed into a system consisting of adaptive, initially unstructured
    feature detectors.  There are also adaptive predictors constantly
    trying to predict current feature detector outputs from other feature
    detector outputs.  Simultaneously, however, the feature detectors try
    to become as unpredictable as possible, resulting in a co-evolution
    of predictors and feature detectors.
    This paper describes the implementation of a visual processing
    system trained by semi-linear predictability minimization, and
    presents many experiments that examine its response to artificial
    and real-world images.  In particular, we observe that under a wide
    variety of conditions, predictability minimization results in the
    development of well-known visual feature detectors.
}}