The Influence Of The Real-Scene Spatial Regularities On The Visual Working Memory For Objects

Visual working memory(VWM)is a cognitive subsystem of working memory with limited capacity and the ability to temporarily store and process information.The realistic scene contains many complex objects,and a large amount of visual information input keeps change constantly.So it is crucial to form relatively stable and accurate object representation in visual working memory for further cognitive operational tasks.Purpose:Previous studies have proved that memory for visual objects in real scenes is based on certain grouping principles,such as functional and semantic associations.In addition,the study about simple stimuli grouping improving VWM performance found decoding and sequence-order effect.Decoding effect means that an individual item within a chunk is not accessible on-line in WM and needs a retrieval process taking time,thereby delaying responses times.The aim of this study was to explore whether objects in real-world scenes can be grouped by general spatial location relationships which led to enhanced visual working memory performance.At the same time,we investigated the rationality of this kind of spatial location grouping by examining the features of processing when a single object was extracted individually.Method:In Study 1 we investigated object pairs with typical,regular spatial relationship but no action relationship.Experiment 1a was a 2(stimuli presentation mode:conforming to spatial location relationship,violating spatial location relationship)X 2(probe stimuli location:upper,lower)factorial design to examine whether there was a decoding process or a sequence-order effect.Experiment 1b was the same as Experiment 1a,except that we did not limit response time in Experiment 1a but we used a response-deadline of 2000ms in experiment 1b to prove the decoding phenomenon in experiment 1a is a general phenomenon.In Study 2 We focused on the object pairs with typical,regular spatial relationship and action relationship in Experiment 2a and 2b.The procedure and design were the same as in Study 1 except the stimuli type.Study 2 tested whether the regulations found in Study 1 would be broken due to the active-passive relationship between the objects.Results:First,in study 1 and study 2,when the objects were presented according to the spatial location relationship of the real scene,the accuracy rate of visual working memory is significantly higher than that when the object is presented in violation of the spatial location relationship,showing the memory advantage brought by the grouping of spatial positions.Second,under the condition of conforming to spatial location relationship it took more time to retrieve a single object and when a pair needed to be decoded,the individual items were retrieved in a stereotyped order from top-object to bottom-object with participants responding quicker on upper objects than on lower objects.Third,when the objects were presented in a way that is consistent with to the spatial location relationship of the real scene,the correct accuracy of the visual working memory of the upper object is significantly higher than that of the lower obj ect.Conclusion:First,the current study suggests that visual working memory performance of objects can be improved by grouping based on the spatial location relationship of the real scene.Second,it was found that there are decoding and sequence-order effect when the object of a pair are represented individually.Third,we found that under the condition of according to spatial location relation,the upper object is not only retrieved first(shorter response time),but also shows greater advantages of visual working memory(higher accuracy rate)compared with the lower object.