Use this URL to cite or link to this record in EThOS:
Title: Decoding attentional load in visual perception : a signal processing approach
Author: Palmer, L. C. T.
ISNI:       0000 0004 7224 3271
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Previous research has established that visual perception tasks high in attentional load (or ‘perceptual load’, defined operationally to include either a larger number of items or a greater perceptual processing demand) result in reduced perceptual sensitivity and cortical response for visual stimuli outside the focus of attention. However, there are three challenges facing the load theory of attention today. The first is to describe a neural mechanism by which load-induced perceptual deficits are explained; the second is to clarify the concept of perceptual load and develop a method for estimating the load induced by a visual task a priori, without recourse to measures of secondary perceptual effects; and the third is to extend the study of attentional load to natural, real-world, visual tasks. In this thesis we employ signal processing and machine learning approaches to address these challenges. In Chapters 3 and 4 it is shown that high perceptual load degrades the perception of orientation by modulating the tuning curves of neural populations in early visual cortex. The combination of tuning curve modulations reported is unique to perceptual load, inducing broadened tuning as well as reductions in tuning amplitude and overall neural activity, and so provides a novel low-level mechanism for behaviourally relevant failures of vision such as inattentional blindness. In Chapter 5, a predictive model of perceptual load during the task of driving is produced. The high variation in perceptual demands during real-world driving allow the construction of a direct fine-scale mapping between high-resolution natural imagery, captured from a driver's point-of-view, and induced perceptual load. The model therefore constitutes the first system able to produce a priori estimates of load directly from visual characteristics of a natural task, extending research into the antecedents of perceptual load beyond the realm of austere laboratory displays. Taken together, the findings of this thesis represent major theoretical advances into both the causes and effects of high perceptual load.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available