<?xml version="1.0" standalone="yes"?>
<Paper uid="W98-1228">
  <Title>Selective Attention and the Acquisition of Spatial Semantics</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> This paper is concerned with the acquisition of natural language spatial semantics by a neurally plausible connectionist system. Within the cognitive framework introduced by (Langacker, 1987), elementary spatial concepts (such as the English above) are characterised by locative relations between a potentially mobile object called the trajector (TR) and a static reference object called the landmark (LM).</Paragraph>
    <Paragraph position="1"> Previous computational investigations of this problem (Regier, 1992), have relied upon highly structured feature detection systems and the abstraction of object identification issues into the input data.</Paragraph>
    <Paragraph position="2"> While highly successful on their own terms, systems of this nature are not readily generalisable to problems involving more sophisticated (especially cluttered) input scenes and linguistic phenomena, and provide neither a conscious nor an autonomous selection mechanism through which such inputs may be successfully processed.</Paragraph>
    <Paragraph position="3"> The model discussed below resolves some of these issues through the use of mechanisms of selective visual attention, through abstraction of established models from computational neuroscience (Niebur and Koch, 1997) and extension to allow linguistic input to cue selection and scene parsing. While retaining the overall computational philosophy of the Berkeley Lo project (see section 2), the present work does not rely upon feature pre-processing to the same extent as the Regier system - representations being based upon probabilistic receptive fields. In this way, 'prior knowledge' of limited specificity may be employed through higher level recruitment to represent quite complex relations (see section 5.1 and (Hogan and Diederich, 1994), (Hogan and Diederich, 1995)).</Paragraph>
    <Paragraph position="4"> The computational philosophy of the Berkeley Lo project is introduced in the next section, followed by discussion of the Regier model and the importance of explicit object recognition in the light of evidence from early language acquisition. Section 2.4 relates this discussion to an accepted cognitive theory mediated through binding of representations at the focus of attention. Selective visual attention, and recent computational models of the process dominate chapter 3, prior to a formal outline of the model in chapter 4. The paper concludes with examination of representations for a limited set of English static concepts - developed through simulations based upon novel Gaussian domain response units - along with discussion of extensions to dynamic concepts.</Paragraph>
  </Section>
class="xml-element"></Paper>