Temporal latencies and position uncertainty in stereoscopic and luminance motion using a continuous eye-tracking task

Poster Presentation 56.428: Tuesday, May 21, 2024, 2:45 – 6:45 pm, Pavilion
Session: Binocular Vision: Disparity, stereopsis and suppression

Ignacio Serrano-Pedraza1 (), Ichasus Llamas-Cornejo1, Omar Bachtoula1; 1Department of Experimental Psychology. Universidad Complutense de Madrid

Stereovision has an unexpected anisotropy: detecting sinusoidal depth corrugations is easier at low spatial frequencies when they are horizontally oriented than when they are vertical. Here, we are interested in the study of the temporal characteristics of this anisotropy. We measured the response of the optomotor system to different suprathreshold stereoscopic stimuli and compared it to the response to luminance stimuli. The stereoscopic stimuli, made with dynamic random dot stereograms, were half-cycle sinusoidal depth corrugations with disparity 2.1 arcsec and the background set to 0 disparity (i.e. 3D bar-shaped on a zero-disparity background). The luminance stimuli were also half-cycle sinusoid with a contrast of 0.8 and with a black background (i.e. white bar on black background). The bar-shaped stimuli moved following a one-dimensional random walk for four seconds and the subject task was to keep the bar centered in the fovea. We tested four spatial frequencies: 0.1, 0.2, 0.4, & 0.8 c/deg and two orientations (e.g. Vertical vs. Horizontal). To analyze the data, we used the eye movement correlogram technique by cross-correlating version velocities and stimulus velocities. The velocity correlogram for luminance stimuli had a positive peak with averaged latency around 140 ms for vertical and 160 ms for horizontal orientation. Interestingly, we found the longer peak latencies for stereo, around 370 ms for horizontal and 425 ms for vertical orientation. We also found higher position uncertainty for stereo than for luminance (using the standard deviation of the positive peak of the correlogram). For luminance stimuli, this uncertainty was higher for horizontal stimuli than vertical ones; and for stereo, the position uncertainty was higher for vertical than horizontal depth corrugations. In summary, our results show longer lags and higher position uncertainty in stereo than in luminance motion perception, and a strong stereoscopic anisotropy with suprathreshold stimuli.

Acknowledgements: Supported by grant PID2021-122245NB-I00 from Ministerio de Ciencia e Innovación (Spain) to ISP.