The role of stereo vision in visual-vestibular integration

John S. Butler, Jennifer L. Campos, Heinrich Bulthoff, Stuart T. Smith

Research output: Contribution to journalArticle

36 Citations (Scopus)

Abstract

Self-motion through an environment stimulates several sensory systems, including the visual system and the vestibular system. Recent work in heading estimation has demonstrated that visual and vestibular cues are typically integrated in a statistically optimal manner, consistent with Maximum Likelihood Estimation predictions. However, there has been some indication that cue integration may be affected by characteristics of the visual stimulus. Therefore, the current experiment evaluated whether presenting optic flow stimuli stereoscopically, or presenting both eyes with the same image (binocularly) affects combined visual-vestibular heading estimates. Participants performed a two-interval forced-choice task in which they were asked which of two presented movements was more rightward. They were presented with either visual cues alone, vestibular cues alone or both cues combined. Measures of reliability were obtained for both binocular and stereoscopic conditions. Group level analyses demonstrated that when stereoscopic information was available there was clear evidence of optimal integration, yet when only binocular information was available weaker evidence of cue integration was observed. Exploratory individual analyses demonstrated that for the stereoscopic condition 90% of participants exhibited optimal integration, whereas for the binocular condition only 60% of participants exhibited results consistent with optimal integration. Overall, these findings suggest that stereo vision may be important for self-motion perception, particularly under combined visual-vestibular conditions.

Original languageEnglish
Pages (from-to)453-470
Number of pages18
JournalSeeing and Perceiving
Volume24
Issue number5
DOIs
Publication statusPublished - 2011 Oct 6

Fingerprint

Stereo vision
Cues
Binoculars
Optic Flow
Maximum likelihood estimation
Motion Perception
Optics
Experiments

Keywords

  • Maximum likelihood estimation
  • Multi-sensory integration
  • Self-motion perception
  • Stereo vision
  • Vestibular

ASJC Scopus subject areas

  • Sensory Systems
  • Cognitive Neuroscience
  • Ophthalmology
  • Experimental and Cognitive Psychology
  • Computer Vision and Pattern Recognition

Cite this

The role of stereo vision in visual-vestibular integration. / Butler, John S.; Campos, Jennifer L.; Bulthoff, Heinrich; Smith, Stuart T.

In: Seeing and Perceiving, Vol. 24, No. 5, 06.10.2011, p. 453-470.

Research output: Contribution to journalArticle

Butler, John S. ; Campos, Jennifer L. ; Bulthoff, Heinrich ; Smith, Stuart T. / The role of stereo vision in visual-vestibular integration. In: Seeing and Perceiving. 2011 ; Vol. 24, No. 5. pp. 453-470.
@article{d6e2095e01a14999ba3f14de5011772c,
title = "The role of stereo vision in visual-vestibular integration",
abstract = "Self-motion through an environment stimulates several sensory systems, including the visual system and the vestibular system. Recent work in heading estimation has demonstrated that visual and vestibular cues are typically integrated in a statistically optimal manner, consistent with Maximum Likelihood Estimation predictions. However, there has been some indication that cue integration may be affected by characteristics of the visual stimulus. Therefore, the current experiment evaluated whether presenting optic flow stimuli stereoscopically, or presenting both eyes with the same image (binocularly) affects combined visual-vestibular heading estimates. Participants performed a two-interval forced-choice task in which they were asked which of two presented movements was more rightward. They were presented with either visual cues alone, vestibular cues alone or both cues combined. Measures of reliability were obtained for both binocular and stereoscopic conditions. Group level analyses demonstrated that when stereoscopic information was available there was clear evidence of optimal integration, yet when only binocular information was available weaker evidence of cue integration was observed. Exploratory individual analyses demonstrated that for the stereoscopic condition 90{\%} of participants exhibited optimal integration, whereas for the binocular condition only 60{\%} of participants exhibited results consistent with optimal integration. Overall, these findings suggest that stereo vision may be important for self-motion perception, particularly under combined visual-vestibular conditions.",
keywords = "Maximum likelihood estimation, Multi-sensory integration, Self-motion perception, Stereo vision, Vestibular",
author = "Butler, {John S.} and Campos, {Jennifer L.} and Heinrich Bulthoff and Smith, {Stuart T.}",
year = "2011",
month = "10",
day = "6",
doi = "10.1163/187847511X588070",
language = "English",
volume = "24",
pages = "453--470",
journal = "Multisensory research",
issn = "2213-4794",
publisher = "Brill",
number = "5",

}

TY - JOUR

T1 - The role of stereo vision in visual-vestibular integration

AU - Butler, John S.

AU - Campos, Jennifer L.

AU - Bulthoff, Heinrich

AU - Smith, Stuart T.

PY - 2011/10/6

Y1 - 2011/10/6

N2 - Self-motion through an environment stimulates several sensory systems, including the visual system and the vestibular system. Recent work in heading estimation has demonstrated that visual and vestibular cues are typically integrated in a statistically optimal manner, consistent with Maximum Likelihood Estimation predictions. However, there has been some indication that cue integration may be affected by characteristics of the visual stimulus. Therefore, the current experiment evaluated whether presenting optic flow stimuli stereoscopically, or presenting both eyes with the same image (binocularly) affects combined visual-vestibular heading estimates. Participants performed a two-interval forced-choice task in which they were asked which of two presented movements was more rightward. They were presented with either visual cues alone, vestibular cues alone or both cues combined. Measures of reliability were obtained for both binocular and stereoscopic conditions. Group level analyses demonstrated that when stereoscopic information was available there was clear evidence of optimal integration, yet when only binocular information was available weaker evidence of cue integration was observed. Exploratory individual analyses demonstrated that for the stereoscopic condition 90% of participants exhibited optimal integration, whereas for the binocular condition only 60% of participants exhibited results consistent with optimal integration. Overall, these findings suggest that stereo vision may be important for self-motion perception, particularly under combined visual-vestibular conditions.

AB - Self-motion through an environment stimulates several sensory systems, including the visual system and the vestibular system. Recent work in heading estimation has demonstrated that visual and vestibular cues are typically integrated in a statistically optimal manner, consistent with Maximum Likelihood Estimation predictions. However, there has been some indication that cue integration may be affected by characteristics of the visual stimulus. Therefore, the current experiment evaluated whether presenting optic flow stimuli stereoscopically, or presenting both eyes with the same image (binocularly) affects combined visual-vestibular heading estimates. Participants performed a two-interval forced-choice task in which they were asked which of two presented movements was more rightward. They were presented with either visual cues alone, vestibular cues alone or both cues combined. Measures of reliability were obtained for both binocular and stereoscopic conditions. Group level analyses demonstrated that when stereoscopic information was available there was clear evidence of optimal integration, yet when only binocular information was available weaker evidence of cue integration was observed. Exploratory individual analyses demonstrated that for the stereoscopic condition 90% of participants exhibited optimal integration, whereas for the binocular condition only 60% of participants exhibited results consistent with optimal integration. Overall, these findings suggest that stereo vision may be important for self-motion perception, particularly under combined visual-vestibular conditions.

KW - Maximum likelihood estimation

KW - Multi-sensory integration

KW - Self-motion perception

KW - Stereo vision

KW - Vestibular

UR - http://www.scopus.com/inward/record.url?scp=80053390011&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80053390011&partnerID=8YFLogxK

U2 - 10.1163/187847511X588070

DO - 10.1163/187847511X588070

M3 - Article

VL - 24

SP - 453

EP - 470

JO - Multisensory research

JF - Multisensory research

SN - 2213-4794

IS - 5

ER -