Conscious access and complexity of visual features

Lindh, Pär Johan Daniel (2022). Conscious access and complexity of visual features. University of Birmingham. Ph.D.

Preview

Lindh2022PhD_Redacted.pdf
Text - Redacted Version
Available under License All rights reserved.
Download (11MB) | Preview

Abstract

Imagine you are in your car driving to meet a friend at a restaurant you
have never been to before. As an experienced driver, you don’t need
to deliberately direct your gaze. Instead, your attention is automatically
drawn to crossings far off in the distance, other moving vehicles, and
relevant road signs. Without having to assert effort, your brain
suppresses details in your immediate surroundings to enhance
relevant information. When you arrive at the restaurant, you swiftly
search through the crowd of strangers, assessing whether everyone is
your friend within a fraction of a second. Your brain effortlessly
evaluates each person with templates in your memory, first on crude
features such as hair colour or height, and for anyone who fits these
criteria, assessment is carried out on finer facial features. With our
ability to use logical inferences based on experience we build
templates of a target, which we use to efficiently scan through our
environment. Regardless of how mundane this everyday task might
seem; its completion requires several fundamental computational
problems to be overcome. When driving a car and when searching a
crowded room, you need to selectively enhance and suppress visual
information, as processing all information equally is an inefficient use
of resources. It can take several hundred milliseconds to fully process
a complex natural scene (Kar et al., 2019), meaning that the
processing of several visual objects must be happening in parallel. To
add to this complexity, humans are continuously updating their goals
(first, search for the bar across the whole room, then search for a
person at the bar) based on information we are gaining within each
moment. In this dissertation, I will address how the brain organizes
information into categories, how items that are processed in parallel
can interfere with each other, and at what levels of processing these
interferences occur.