---
title: "Learning Shape Descriptions"
authors:
  - "Jonathan Connell"
  - "Michael Brady"
doi: ""
url: "https://ijcai.org/proceedings/1985/186"
pdf: "https://ijcai.org/proceedings/1985/0186.pdf"
venue: "Proceedings of the Ninth International Joint Conference on Artificial Intelligence"
year: 1985
pages: "922-924"
track: "Perception"
keywords:
---

# Learning Shape Descriptions

Jonathan Connell, Michael Brady

*Proceedings of the Ninth International Joint Conference on Artificial Intelligence*, pages 922-924.

## Abstract

We report on initial experiments with an implemented learning system whose inputs are images of two-dimensional shapes. The system first builds semantic network shape descriptions bascd on Brady's smoothed local symmetry representation. lt learns shape models from them using a modified version of Winston's ANALOGY progranl. The learning program uses only positive examples, and is capab le of learning disjunctive concepts. We discuss the learnability of shape descriptions.