Have there been any studies done that test the reaction time to vision processing tasks with and without the benefit of binocular disparities? I have been wondering how much depth information (such as the jutting out of the nose on a face) contributes to visual processing and object encoding in the brain - or if the visual scene is treated like a diorama with 2D images placed at various depths.

My assumption is that if the 3D aspect of the scene components are important then the reaction times would vary.

  • $\begingroup$ Welcome! Interesting question. To make it more concrete, could you include some example vision processing tasks you know about already, and are thinking of you would want to test? $\endgroup$
    – Steven Jeuris
    Jan 5 '19 at 17:11
  • 1
    $\begingroup$ Thanks, well I suppose I was just thinking of some sort of random object recognition tasks - I'm coming to this from a programming perspective, have been studying deep learning networks all of which i'm aware of just use 2D images so I'm pondering whats to be gained by adding an extra depth channel to a ConvNet and whether that's conceivably whats all ready included in the biological case. $\endgroup$
    – norlesh
    Jan 5 '19 at 17:24
  • $\begingroup$ - the task would somehow time responses to displayed 3D items revealed behind flat glass vs displayed on a flat screen display perhaps. $\endgroup$
    – norlesh
    Jan 5 '19 at 17:50

According to these three published studies I found below - reaction time is slower when depth disparity information is withheld, as to whether or not that extra information is actually encoded in the brain is still an open question.

  • This first paper by Young Lim Lee; Jeffrey A. Saunders used left and right eye rendered images of novel 3D shapes with and without binocular cues to time test subjects in discriminating whether two images were rotational views of the same object. With a set of three experiments they concluded in all cases that the stereo information improved reaction times.

    Stereo improves 3D shape discrimination even when rich monocular shape cues are available

  • The second paper by Darren Burke, Jessica Taubert, Talia Higman had test subjects view pairs of photographed human faces briefly through a stereoscope with and without stereoscopic information and were required to identify if it was the same or a different individual. Results were that the subjects had faster reaction times and lower error rates when stereoscopic information was present.

    Are face representations viewpoint dependent? A stereo advantage for generalising across different views of faces

  • And the third paper by Oliver ZJ, Cristino F, Roberts MV, Pegna AJ, Leek EC the test subjects were split into a mono and stereo groups before being required to memomrize a set of 3D objects and then determine amongst a series of test objects whether or not it was from the set. The test was conducted using a 3D stereo monitor and glasses with the images displayed either with complementary or identical images to either eye depending on the group while 128 channel ERP (event-related-potential) traces were recorded. Results showed higher accuracy for the stereo group and while there was no statistics given for reaction times, the paper concludes with an analysis and discussion regarding the differences in ERP data collected from the two groups.

    Stereo Viewing Modulates Three-Dimensional Shape Processing During Object Recognition: A High-Density ERP Study


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.