Dynamics of bi-stable perception of plaids

J. M. Hupé, N. Rubin

Research output: Contribution to journalArticle

Abstract

Background. The time spent perceiving each interpretation in ambiguous displays has been used as a measure of their relative strength (e.g., Levelt 1965 for binocular rivalry; Von Grunau and Dubé 1993 for plaids). At ARVO 2000 we introduced another dynamical measure: the delay between stimulus presentation and the first report of transparent motion (sliding), given by observers' reaction time to see sliding, RTtransp. We showed that RTtransp could be manipulated by form cues that affect the strength of the sliding interpretation: RTtransp was longer for luminance configurations inconsistent with transparency (Stoner et al. 1990). Purpose. What is the relationship between these two dynamical variables, the overall time spent seeing transparent (T) versus coherent (C) motion, and RTtransp? Methods. Plaid stimuli (Wallach 1935) were displayed for (RTtransp+60) sec. Observers held down a mouse button as long as they perceived motion transparency and released it when they perceived coherency. We measured RTtransp, T and C (=60-T), and the average duration of transparency and coherence episodes, Tav and Cav. Plaid parameters (direction of motion, angle between gratings, speed, duty cycle) were varied. Results. RTtransp is closely correlated to both C and Cav. All three variables co-vary for all the parameters tested. Interestingly, the relative strengths of transparency vs. coherency as revealed by RTtransp may be quite different than estimates obtained from short-exposure 2AFC procedures, which are more susceptible to 'ceiling effects' (e.g., for stimuli where C=T, or Cav=Tav, sliding is typically seen only after RTtransp =1̃4s). In addition, RTtransp can reveal effects of relevant parameters even in regimes where only sliding is reported (C=0). Conclusion. 1- RTtransp is a time-efficient and invaluable dependent variable to study bi-stable percepts. 2- The relations discovered between RTtransp, C and Cav impose new constraints on models of bi-stability.

Original languageEnglish (US)
JournalJournal of Vision
Volume1
Issue number3
DOIs
StatePublished - 2001

Fingerprint

Reaction Time
Cues
Direction compound

ASJC Scopus subject areas

  • Ophthalmology

Cite this

Dynamics of bi-stable perception of plaids. / Hupé, J. M.; Rubin, N.

In: Journal of Vision, Vol. 1, No. 3, 2001.

Research output: Contribution to journalArticle

Hupé, J. M. ; Rubin, N. / Dynamics of bi-stable perception of plaids. In: Journal of Vision. 2001 ; Vol. 1, No. 3.
@article{aae95ab9bc1749eb98b190eada62b327,
title = "Dynamics of bi-stable perception of plaids",
abstract = "Background. The time spent perceiving each interpretation in ambiguous displays has been used as a measure of their relative strength (e.g., Levelt 1965 for binocular rivalry; Von Grunau and Dub{\'e} 1993 for plaids). At ARVO 2000 we introduced another dynamical measure: the delay between stimulus presentation and the first report of transparent motion (sliding), given by observers' reaction time to see sliding, RTtransp. We showed that RTtransp could be manipulated by form cues that affect the strength of the sliding interpretation: RTtransp was longer for luminance configurations inconsistent with transparency (Stoner et al. 1990). Purpose. What is the relationship between these two dynamical variables, the overall time spent seeing transparent (T) versus coherent (C) motion, and RTtransp? Methods. Plaid stimuli (Wallach 1935) were displayed for (RTtransp+60) sec. Observers held down a mouse button as long as they perceived motion transparency and released it when they perceived coherency. We measured RTtransp, T and C (=60-T), and the average duration of transparency and coherence episodes, Tav and Cav. Plaid parameters (direction of motion, angle between gratings, speed, duty cycle) were varied. Results. RTtransp is closely correlated to both C and Cav. All three variables co-vary for all the parameters tested. Interestingly, the relative strengths of transparency vs. coherency as revealed by RTtransp may be quite different than estimates obtained from short-exposure 2AFC procedures, which are more susceptible to 'ceiling effects' (e.g., for stimuli where C=T, or Cav=Tav, sliding is typically seen only after RTtransp =1̃4s). In addition, RTtransp can reveal effects of relevant parameters even in regimes where only sliding is reported (C=0). Conclusion. 1- RTtransp is a time-efficient and invaluable dependent variable to study bi-stable percepts. 2- The relations discovered between RTtransp, C and Cav impose new constraints on models of bi-stability.",
author = "Hup{\'e}, {J. M.} and N. Rubin",
year = "2001",
doi = "10.1167/1.3.157",
language = "English (US)",
volume = "1",
journal = "Journal of Vision",
issn = "1534-7362",
publisher = "Association for Research in Vision and Ophthalmology Inc.",
number = "3",

}

TY - JOUR

T1 - Dynamics of bi-stable perception of plaids

AU - Hupé, J. M.

AU - Rubin, N.

PY - 2001

Y1 - 2001

N2 - Background. The time spent perceiving each interpretation in ambiguous displays has been used as a measure of their relative strength (e.g., Levelt 1965 for binocular rivalry; Von Grunau and Dubé 1993 for plaids). At ARVO 2000 we introduced another dynamical measure: the delay between stimulus presentation and the first report of transparent motion (sliding), given by observers' reaction time to see sliding, RTtransp. We showed that RTtransp could be manipulated by form cues that affect the strength of the sliding interpretation: RTtransp was longer for luminance configurations inconsistent with transparency (Stoner et al. 1990). Purpose. What is the relationship between these two dynamical variables, the overall time spent seeing transparent (T) versus coherent (C) motion, and RTtransp? Methods. Plaid stimuli (Wallach 1935) were displayed for (RTtransp+60) sec. Observers held down a mouse button as long as they perceived motion transparency and released it when they perceived coherency. We measured RTtransp, T and C (=60-T), and the average duration of transparency and coherence episodes, Tav and Cav. Plaid parameters (direction of motion, angle between gratings, speed, duty cycle) were varied. Results. RTtransp is closely correlated to both C and Cav. All three variables co-vary for all the parameters tested. Interestingly, the relative strengths of transparency vs. coherency as revealed by RTtransp may be quite different than estimates obtained from short-exposure 2AFC procedures, which are more susceptible to 'ceiling effects' (e.g., for stimuli where C=T, or Cav=Tav, sliding is typically seen only after RTtransp =1̃4s). In addition, RTtransp can reveal effects of relevant parameters even in regimes where only sliding is reported (C=0). Conclusion. 1- RTtransp is a time-efficient and invaluable dependent variable to study bi-stable percepts. 2- The relations discovered between RTtransp, C and Cav impose new constraints on models of bi-stability.

AB - Background. The time spent perceiving each interpretation in ambiguous displays has been used as a measure of their relative strength (e.g., Levelt 1965 for binocular rivalry; Von Grunau and Dubé 1993 for plaids). At ARVO 2000 we introduced another dynamical measure: the delay between stimulus presentation and the first report of transparent motion (sliding), given by observers' reaction time to see sliding, RTtransp. We showed that RTtransp could be manipulated by form cues that affect the strength of the sliding interpretation: RTtransp was longer for luminance configurations inconsistent with transparency (Stoner et al. 1990). Purpose. What is the relationship between these two dynamical variables, the overall time spent seeing transparent (T) versus coherent (C) motion, and RTtransp? Methods. Plaid stimuli (Wallach 1935) were displayed for (RTtransp+60) sec. Observers held down a mouse button as long as they perceived motion transparency and released it when they perceived coherency. We measured RTtransp, T and C (=60-T), and the average duration of transparency and coherence episodes, Tav and Cav. Plaid parameters (direction of motion, angle between gratings, speed, duty cycle) were varied. Results. RTtransp is closely correlated to both C and Cav. All three variables co-vary for all the parameters tested. Interestingly, the relative strengths of transparency vs. coherency as revealed by RTtransp may be quite different than estimates obtained from short-exposure 2AFC procedures, which are more susceptible to 'ceiling effects' (e.g., for stimuli where C=T, or Cav=Tav, sliding is typically seen only after RTtransp =1̃4s). In addition, RTtransp can reveal effects of relevant parameters even in regimes where only sliding is reported (C=0). Conclusion. 1- RTtransp is a time-efficient and invaluable dependent variable to study bi-stable percepts. 2- The relations discovered between RTtransp, C and Cav impose new constraints on models of bi-stability.

UR - http://www.scopus.com/inward/record.url?scp=4143092780&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4143092780&partnerID=8YFLogxK

U2 - 10.1167/1.3.157

DO - 10.1167/1.3.157

M3 - Article

VL - 1

JO - Journal of Vision

JF - Journal of Vision

SN - 1534-7362

IS - 3

ER -