Abstract
We seldom mistake a closer object as being larger, even though its retinal image is bigger. One underlying mechanism could be to calculate the size of the retinal image relative to that of another nearby object. Here we set out to investigate whether single neurons in the monkey inferotemporal cortex (IT) are sensitive to the relative size of parts in a display. Each neuron was tested on shapes containing two parts that could be conjoined or spatially separated. Each shape was presented in four versions created by combining the two parts at each of two possible sizes. In this design, neurons sensitive to the absolute size of parts would show the greatest response modulation when both parts are scaled up, whereas neurons encoding relative size would show similar responses. Our main findings are that 1) IT neurons responded similarly to all four versions of a shape, but tuning tended to be more consistent between versions with proportionately scaled parts; 2) in a subpopulation of cells, we observed interactions that resulted in similar responses to proportionately scaled parts; 3) these interactions developed together with sensitivity to absolute size for objects with conjoined parts but developed slightly later for objects with spatially separate parts. Taken together, our results demonstrate for the first time that there is a subpopulation of neurons in IT that encodes the relative size of parts in a display, forming a potential neural substrate for size constancy.
Highlights
BECAUSE THE VERY ACT of image formation destroys depth, many accounts of size constancy require additional knowledge about the external world
To investigate whether neurons encode the relative size of parts, we performed an analysis of variance (ANOVA) on the firing rate evoked by the four versions of each shape, with the sizes of the large (1x/2x) and small (1x/2x) part as factors
The main finding of this study is that a subpopulation of neurons in monkey inferotemporal cortex (IT) encodes the relative size of parts in a display
Summary
BECAUSE THE VERY ACT of image formation destroys depth, many accounts of size constancy require additional knowledge about the external world. Each part was presented at two possible sizes (1x or 2x), resulting in a total of four versions per object (see Fig. 2) Throughout, we denote these versions as v11, v12, v21, and v22 (where, e.g., v12 represents the version with the small part at size 1x and the large part at size 2x). It would respond more to v11 and v22—which represent both parts scaling 2x in size—than to other versions such as v12 and v21 where the parts scale disproportionately This would produce an interaction effect in the same ANOVA. Our results show that while the majority of IT neurons are modulated by the absolute size of parts, a subpopulation of neurons encodes the relative size of parts in a display, forming a potential neural substrate for size constancy
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
