Selective Attention Model of Moving Objects

Summary

---

Tracking moving objects is a vital visual task for the survival of an animal. We describe oscillatory neural network models of visual attention with a central element that can track a moving target among a set of distracters on the screen. At the initial stage, the model forms the focus of attention on an arbitrary object that is considered as a target. Other objects are treated as distracters. We present here two models: 1) synchronisation based model designed as a network of phase oscillators and 2) spiking neural model which is based on the idea of resource-limited parallel visual pointers. Selective attention and the tracking process are represented by the partial synchronisation between the central unit and a subgroup of peripheral elements. Simulation results are in overall agreement with the findings from psychological experiments: overlapping between the target and distractors is the main source of errors.

See the full content of this document

Extract

---

Selective Attention Model of Moving Objects

(ProQuest: ... denotes formulae omitted.)

1. Introduction

Selective visual attention is a mechanism that gives a living organism the possibility to extract from the incoming visual information a part that is most important at a given moment and that should be processed in more detail. This mechanism is necessary due to limited processing capabilities of the visual system which does not allow rapid analysis of the whole visual scene.

The important property of attention is its metastability. This means that after being fixed, the focus of attention does not change for some time even when objects in a scene gradually vary their parameters (shape, brightness, position). In particular, the metastability of attention makes it possible to track a moving object. Special conditions should be fulfilled for attention to be switched from one object to another. These conditions include (a) an abrupt change of parameters of an object in the focus of attention, (b) appearance and disappearance of objects in the scene, (c) overlapping or hiding of objects due to their movements, (d) termination or voluntary break of object processing.

In recent years attention has become a popular field for neural network modelling. The models of attention can be subdivided into two categories. Connectionist models (for example, [1-4]) are based on a winner-takes-all procedure and are implemented through a proper modification of the weights of connections in a hierarchical neural network...

See the full content of this document

---