Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

TO INFEROTEMPORAL NEURONS THE WHOLE IS NOT THE SUM OF THE PARTS

Hare, Erin (2017) TO INFEROTEMPORAL NEURONS THE WHOLE IS NOT THE SUM OF THE PARTS. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

This is the latest version of this item.

[img] PDF
Restricted to University of Pittsburgh users only until 25 June 2018.

Download (3MB) | Request a Copy

Abstract

Vision seems to occur effortlessly and without mistakes. As a result, it is easy to lose sight of the complex representational mechanisms going on under the hood. In macaque monkeys, the brain region thought to be the ultimate mediator of object recognition is the inferotemporal cortex (IT). The purpose of this dissertation was to investigate how IT neurons handle parts of objects, both in a context wherein parts interfere with one another as well as a context in which certain parts aren’t readily perceived. We broke this down into three distinct projects, outlined below.

The first project was concerned with the behavioral phenomenon known as crowding, in which clutter causes peripheral objects to devolve into an unintelligible jumble. We are the first to develop a task that was conducive to concurrent behavior and neuronal recordings in monkeys. To demonstrate relevance of our task, we turned to a hallmark of crowding: that what matters is the eccentricity and spacing between objects, not object size. Having demonstrated this, we were set to proceed to neuronal recordings.

Our primary question was whether crowding quantitatively reduced the strength of IT neuronal selectivity or alternatively whether crowding induced a qualitative change to the neuronal code. Our results support the latter hypothesis. We then asked additional follow-up questions regarding size-sensitivity and adjacency of part-part interactions. Overall, our results were incompatible with a pooling model of crowding and consistent with models based on attention, texture, or source confusion.

The final experiment was concerned with whether certain parts of compound objects were preferentially represented over others. To do this we recorded IT spiking activity while monkeys viewed composite shapes made up of overlapping outlines, as well as all the possible constituent closed parts created by the overlap. Humans tend to only perceive the simpler shapes originally used to create the composite, but the same was not true of IT neurons. Instead, they represented the composite more like its external contour than any other part, especially in the initial phase of the response.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Hare, Erinecrowder@pitt.eduecrowder
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee MemberOlson, Carlcolson@cnbc.cmu.edu
Committee ChairBehrmann, Marlenebehrmann@cmu.edu
Committee MemberCohen, Marlenecohenm@pitt.educohenm
Committee MemberSmith, Mattsmithma@pitt.edusmithma
Committee MemberLee, Tai Singtai@cnbc.cmu.edu
Committee MemberSheinberg, Daviddavid_sheinberg@brown.edu
Date: 25 June 2017
Date Type: Publication
Defense Date: 6 April 2017
Approval Date: 25 June 2017
Submission Date: 19 December 2016
Access Restriction: 1 year -- Restrict access to University of Pittsburgh for a period of 1 year.
Number of Pages: 147
Institution: University of Pittsburgh
Schools and Programs: Dietrich School of Arts and Sciences > Neuroscience
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: monkey, vision, inferotemporal, attention, crowding, gestalt
Date Deposited: 25 Jun 2017 20:39
Last Modified: 25 Jun 2017 20:39
URI: http://d-scholarship.pitt.edu/id/eprint/31423

Available Versions of this Item

  • TO INFEROTEMPORAL NEURONS THE WHOLE IS NOT THE SUM OF THE PARTS. (deposited 25 Jun 2017 20:39) [Currently Displayed]

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item