A unified model for visual masking and crowding.

Visual masking is a tool that is widely used to study information processing in the brain and refers to impaired performance on a target stimulus when a mask stimulus is briefly presented before, during, or after the target, at the same, or at flanking locations. Crowding refers to impaired ability to recognize object in clutter. Though crowding and masking share very similar features, usually they are seen as different phenomena with separate mechanisms and models, which are not yet understood. We predict that masking and crowding are the highest inside the receptive field (RF) and decrease with increasing distance from the RF. We used our recently developed method for estimating the size of the RF. The method is based on a Yes/No detection task, measuring the falsepositive reports (false-alarm, FA) and the hits rate (Phit) for a low-contrast Gabor target placed between two flankers. The target-flanker distance, flanker orientation and target contrast were manipulated. The stimuli appeared randomly either in the fovea or in the periphery (4 deg). We measured the masking (spatial and temporal) and crowding effects at these locations. As predicted, our results show that the masking and crowding effects are correlated with the estimated size of the RF. Inside the RF, strong crowding and 95 masking effects were found, effects that decreased outside the RF. Since the RF is larger in the periphery, these effects are larger in the periphery, but when compensating for the RF size, the effects are similar in the fovea and in the periphery. We conclude that masking and crowding share common mechanisms and both depend on the size of the RF. Thus, a unified low-level model of masking may explain the spatial and temporal masking effect as well as the crowding effect.