A review of the visual masking literature in the context of known neurophysiological and psychophysical properties of the visual system's spatiotemporal response reveals that three consistent and typical pattern masking effects-(a) Type B forward or paracontrast, (b) Type B backward or metacontrast, and (c) Type A forward and backward-can be explained in terms of three simple sensory processes. It is hypothesized that sustained channels are involved in the processing of structural or figural information, whereas transient channels are involved in signaling the spatial location or change in spatial location (motion) of a stimulus. In the proposed model, Type B forward masking or paracontrast is mediated via lateral inhibition realized in the center-surround antagonism of the receptive fields of the sustained cells. In apposition to this mechanism of z'ntrachannel inhibition, Type B backward masking or metacontrast is produced by a mechanism of z'nterchannel inhibition, that is, transient cells laterally inhibit the activity of sustained cells. Both mechanisms are assumed to be operating at or prior to the contour-forming levels of visual processing. Type A masking effects are explained in terms of sensory integration of sustained channel information at preiconic and iconic levels of visual processing. The implications of this multichannel model for saccadic-suppression and information-processing approaches to pattern recognition are discussed.The visibility of a briefly displayed target another briefly displayed masking pattern, pattern can be reduced in several ways by Perhaps the most effective way is to display concurrently with the target a spatially over-~~ lapping pattern mask of sufficiently high con-This study was supported in part by National Eye trast whicn part ially or totally obscures the Institute Grant EY01241-02 to the second author.. . TT u -j ^-J.-L. t, • Much of the scholarly and experimental research target. However, besides this rather obvious that contributed to this paper was done while the type of masking procedure, methods of visual first author was on leave of absence from the Uni-pattern masking have been employed in which versity of Houston, between July 1973 and August t h e target and mask displays do not overlap 1974, at the Bell Laboratories, Murray Hill New . j d/ g h fad Jersey. He thanks Bela Julesz of the Bell Labora-. '•" tories for sponsoring and encouraging his research especially interesting because they can be efforts. We also thank Naomi Weisstein for helpful employed to probe the spatiotemporal propcriticisms and suggestions.erties of information processing and, in par-B^K/S^SS^oio^SiSlSS ticular > Pattern-forming operations in the Houston, Houston, Texas 77004. visual system.

Event-related brain potentials (ERPs) in response to tachistoscopically presented photographs of 2 human faces were recorded for 4--7-month-old infants. For each infant 1 face was chosen to be presented frequently (p = .88, a low-information event) and the other infrequently (p = .12, a high-information event). Both types of events elicited in our infants a long-latency negative ERP wave (ca. 700 msec), termed Nc, and a long-latency positive wave (ca. 1,360 msec), termed Pc. We found that the discrepant, infrequently presented face elicited Nc waves which were higher in amplitude and longer in latency than those elicited by the frequent face. These differences suggest that our infants were able to remember the frequently presented face from trial to trial and to discriminate it from the discrepant face. The discrepant event elicited Pc waves which were insignificantly higher in amplitude than those elicited by frequent events. In adults and children, discrepant events have been found by numerous researchers to elicit positive P3 waves (latency ca. 300--800 msec). In our study, however, such waves could not be discerned. So, of all of the ERP waves which have been related to cognitive processes, the wave which is maturationally the earliest to appear is the Nc wave, which has been related to the perception of attention-getting events or events of interest to the subject. Our findings suggest that ERP responses could provide a sensitive means for investigating infant cognitive development since they do not depend upon an integrated motor-response system.

No abstract

No abstract

Prolonged inspection of uniformly moving contours affects differentially the luminance threshold for the detection of test contours as a function of the direction of motion of the test contours. This finding supports a new explanation of the well-known aftereffect.