Abstract
Purpose. The parameter E2 in the equation F=1+E/E2 is used to characterize the rate at which stimulus size must increase with eccentricity (E) to achieve foveal levels of performance in detection and discrimination tasks. Most previous research has employed 2D stimuli for which discrimination can be accomplished with V1 mechanisms. Here we ask whether size scaling is sufficient to equate the perception of 3D shape-from-texture and shape-from-motion tasks across the visual field.
Method. Both tasks employed 3D surfaces comprising hills, valleys and plains in three possible locations. Therefore there were 27 different surfaces, yielding a 27 alternative forced choice task. Surface shape was conveyed by texture or relative motion. Subjects performed the task at eccentricities of 0 to 16 degrees in the right visual field over a 64 fold range of stimulus sizes.
Results. For both tasks performance reached 100% correct at all eccentricities. Data for each subject were fit with a Gaussian integral at each eccentricity and from these fits E2 values were recovered. For the shape-from-texture task the average E2 was 1.52 and for the shape-from-motion task the average E2 was .61.
Conclusions. In all cases, scaling with F=1+E/E2 eliminated all eccentricity variation from the data. The E2s differed substantially but both average values imply cortical limitations; different limitations may be involved in the two tasks. The data show clear evidence that size scaling is sufficient to equate the perception of shape-from-texture and shape-from-motion across the visual field; at least for the class of stimuli studied here.