Optométrie : accès libre

Abstrait

Stereoscopic Memory When Stimuli No Longer Persist: Void and Binocular Intervals in Alternating Monocular Presentations

Jacques Ninio and Svetlana Rychkova

Objective: Studying some temporal aspects of stereoscopic processing in order to clarify phenomena of stereoscopic persistence and investigate how the brain deals with stereoscopic stimuli that involve various types of clues or difficulties (simple versus complex, with or without oriented elements, with or without curvature, with explicit or camouflaged shapes, with or without disparity discontinuities).
Method: Left and right images of 20 different stereo pairs were generated, one above the other, on a laptop computer display and were observed by the left and right eye of each subject separately wearing proper prism glasses. The stimuli were presented to six subjects under alternating presentation conditions (i) under pure alternation (ii) with an intercalated variable void interval V between the monocular left and right stimulus intervals (iii) with a variable temporal overlap of the two monocular stimuli, thus producing a binocular interval B between the purely monocular intervals. For each V, B, or null interval we determined the longest duration M of the monocular presentations that was compatible with stereopsis.
Results: A modest increase in the duration of binocular intervals B makes it possible much larger increase in the duration of the purely monocular presentation M. The M’s are related to the binocular intervals B’s by a power law, of exponent around 1.3 - 2.5 depending on stimulus complexity. At large B’s, when M is increased, the subject experiences a transition from stable to pulsating stereopsis, then a transition from pulsating to no stereopsis. In an intermediate range of B values, the representative curve splits in 2 branches, an upper one separating stable or pulsating stereopsis from no stereopsis, and a lower branch separating stable from pulsating stereopsis. At some point, the subject reports stable stereopsis, but would be in a regime of “pulsating stereopsis” of which he/she is not aware. The results are represented according to a general phase diagram containing 4 phases. A pair of closely related stimuli can give rise, in some subjects, to strikingly different behaviours at large B’s.
Conclusion: We suggest the existence of a rather stable stereoscopic memory store, allowing a stable 3d interpretation of scenes. Various types of data would decay at different rates and stereoscopic processing, sustained over several alternations, would require brief occasional updates.