The present project mainly addressed the following three questions: 1) Are there separate or bounded features stored in visual working memory (VWM)? 2) What is the role of attention for the sustain of bounded features in VWM? 3) How can human vision system bind features of different time coures into a coherent percept? Results of our studies showed that whether VWM performance would be impaired by a sendary Multiple Object Tracking (MOT) task depended on its content. VWM for color-location combination would be impaired but VWM for color-shape combination would not. When participants were asked to remember the colors of the presented patches, locations of the patches, irrelevant to the task, were processed implicitly. Furthermore, when lables were attached to objects in an MOT task, tracking performance would be increased and digital lables were more effective than color lables. Besides, outcome of objective algorithms based on global and local features were found to be more consistent with human performance in a face recognition task. The above results suggest that it is likely for humen to store and process the object features in a bounded way and attention plays an important role inside which assigns different weights to different dimension of features according to the task requirments. And the results of our
修改评论