Use of Machine Learning to Generate Rules

@inproceedings{Hutber1987UseOM,
  title={Use of Machine Learning to Generate Rules},
  author={David Hutber and P. F. Sims},
  booktitle={Alvey Vision Conference},
  year={1987}
}
The use of Machine Learning techniques applied to visual data is described, within the context of an Alvey exemplar to detect cars in outdoor scenes. A Similarity-Based Learning scheme is employed, that uses segmented images containing cars, to produce rules which are subsequently able to label unknown images. The method is shown to be useful in developing a suitable Knowledge Representation for this vision problem. 

Figures from this paper

Image Labelling with a Neural Network

The lack of contextual integrity in region labelling schemes for segmented visual images has been a long standing problem in computer vision and so far no general regionlabelling system has been found.

Alvey MMI-007 Vehicle Exemplar : Performance & Limitations

This work has adopted a knowledge-based approach to the detection and identification of vehicles in natural scenes, which seeks to explain the 2-D image in terms of known context and known geometry of the 3-D object being identified.

Alvey MMI-007 Vehicle Exemplar: Evaluation and Verification of Model Instances

Using a model-based hypothesise-and-lcst strategy, hypothesised two-dimensional instances of threedimensional models can be used to match against directly observable, view-specific image features, which obviates the necessity of obtaining three-dimensional data from the image.

Alvey MMI-007 Vehicle Exemplar: The Knowledge Based Approach

The central task confronting practical vision systems is to resolve the distinction between data-driven methods and goal-directed methods, and interpret the image data in terms of object abstractions.

Alvey MMI-007 Vehicle Exemplar: Image Segmentation and Attribute Generation

The MMI-007 project uses both bottom-up and top-down strategies to deal with the difficult problem of object recognition in outdoor scenes, and calls upon a range of static segmentation techniques, each individually capable of capturing some aspect of the diffuse information present in the images.

References

SHOWING 1-9 OF 9 REFERENCES

Learning Shape Descriptions

An implemented learning system whose inputs are images of two-dimensional shapes is implemented, which first builds semantic network shape descriptions based on Brady's smoothed local symmetry representation, and is capable of learning disjunctive concepts.

Artificial Intelligence

The history, the major landmarks, and some of the controversies in each of these twelve topics are discussed, as well as some predictions about the course of future research.

Support Logic Programming

A support logic programming system which uses a theory of support pairs to model various forms of uncertainty and uses ideas from Zadeh's fuzzy set theory and Shafer's evidence theory for knowledge representation and inference suitable for expert systems is described.

A.Page 'A Segmentation Algorithm for Colour

  • A.Page 'A Segmentation Algorithm for Colour

Image Segmentation and Attribute Generation Alvey Contract 'The Identification of Objects in 2D Images

  • Proc. AVC
  • 1987

Object Hypothesis by Evidential Reasoning', University of Bristol, Alvey Contract 'The Identification of Objects in 2D Images

  • Proc. AVC
  • 1987

Sims ' Learning Rules from Uncertain Data ' Alvey Contract ' The Identification of Objects in 2 D Images ' Alvey Ref

  • Brisdon ' Instance Verification and Refinement ' , University of Reading , Alvey Contract ' The Identification of Objects in 2 D Images ' , Proc . AVC
  • 1987

Instance Verification and Refinement' , University of Reading, Alvey Contract 'The Identification of Objects in 2D Images

  • Proc. AVC
  • 1987

Learning Rules from Uncertain Data' Alvey Contract 'The Identification of Objects in 2D Images' Alvey Ref

  • Learning Rules from Uncertain Data' Alvey Contract 'The Identification of Objects in 2D Images' Alvey Ref
  • 1986