Abstract
Introduction: Unlike many visual tasks, contrast detection and discrimination are reportedly unchanged by practice (Dorais & Sagi, 97; Adini et al, 02). We re-visited this issue and found very different data.
Methods: The stimulus was a Gabor patch presented either at the visual fovea (6 cpd) or at 5 deg temporal or nasal periphery (0.75 cpd). Contrast thresholds were measured with a 2AFC staircase method.
Results: (1) Among five observers who practiced foveal contrast detection and discrimination at 0, 0.3, 0.47, and 0.63 contrasts, four showed significant improvement over 4–5 days (2 hrs/day). The fifth observer started with very low thresholds and practice did not further improve performance. (2) Practice induced improvement for discriminating a vertical, 0.47 contrast foveal Gabor only partially transfers to discrimination of a horizontal Gabor, for Gabors with spatial frequency +/−1 octave away, and for Gabors at lower (0.30) and higher (0.74) contrasts, with other stimulus conditions identical. (3) Practice also improved contrast discrimination in the visual periphery. Improvement in one hemi-field only partially transfers to the other hemi-field of the same eye. (4) Practice produced much less learning in detection than in high-contrast discrimination under comparable practice conditions. Practicing detection with additional Gabor flankers did not facilitate learning in detection with no flankers when both conditions were run together.
Conclusions: (1) Practice improves contrast discrimination in both foveal and peripheral vision. (2) Learning is partially general, partially specific to the stimulus spatial frequency, orientation, contrast, and retinal location, indicating that both low-level (gain) and high-level (strategy) learning is involved. (3) Contextual stimuli do not affect perceptual learning of contrast detection.