"Seeing where" happens earlier and faster than "seeing what“
Tue-B16-Talk IV-03
Presented by: Christian H. Poth
Recent evidence suggests that visual processing for perception within the ventral visual “what” stream not only computes feature (e.g., color) and category information (e.g., object identity) but also spatial information (e.g., object position). Thus, an object’s category and its location at least partly seem to be processed by the same mechanisms. Therefore, we ask whether the location of an object and its object category are processed equally or whether they differ in how fast they are processed and in how early their processing starts. To this end, participants performed a letter report task in which they briefly viewed a letter that was terminated by a mask display. Afterwards, they reported the identity of the letter (alphanumeric category) and its spatial location. Assessing report performance as a function of presentation duration of the letter and using the Theory of Visual Attention (TVA, Bundesen, 1990, Dyrholm et al., 2011), we obtained measures for the onset (temporal threshold of conscious perception) and the speed of visual processing for both features, letter location and letter identity. We found for location reports that the visual processing speed was higher than for reports of letter identity and, in contrast to a key assumption of TVA, that the temporal threshold of conscious perception was lower. Thus, seeing where an object was happened earlier and faster than seeing what the object was.
Keywords: visual attention, object recognition, TVA