Search

Data sets (Experiments 1 and 2) and R code to perform event history analysis

To study the time course of the global precedence, global-to-local interference, and local-to-global interference effects in a global/local matching task, we analyzed the response time (RT) and accuracy distributions using discrete-time event history analysis and conditional accuracy analysis, respectively. In the matching task, participants saw two compound shapes on either side of a screen (e.g., a circle composed of triangles on the left and a square composed of circles on the right) and were asked to indicate whether the shapes were the same or different on a global (overall shape) or local (smaller shapes) level. A global precedence effect (faster response times in the global task) was found to occur between 300 and 800 ms after stimulus onset. Local-to-global interference was found between 400 and 700 ms. Global-to-local interference was evident by early errors when the local shapes differed, and by a decreased hazard probability of response occurrence between 400 and 800 ms. These results suggest that performance depends on the dynamic interplay between three types of shape comparisons or response tendencies: between two local shapes of the same global stimulus (triggering always the "same" response), between two global shapes ("same" or "different"), and between two local shapes on the left and right of fixation ("same" or "different").

Multiple accounts have been proposed to explain category-specific recognition impairments. Some suggest that category-specific deficits may be caused by a deficit in recurrent processing between the levels of a hierarchically organized visual object recognition system. Here, we tested predictions of interactive processing theories on the emergence of category-selective naming deficits in neurologically intact observers and in patient GA, a single case showing a category-specific impairment for natural objects after a herpes simplex encephalitis infection. Fragmented object outlines were repeatedly presented until correct naming occurred (maximum 10 times), and the fragments increased in length with every repetition. We studied how shape complexity, object category, and fragment curvature influence the timing of correct object identification. The results of a survival analysis are consistent with the idea that deficits in recurrent processing between low- and high-level visual object representations can cause category-selective impairments. ispartof: Visual Cognition vol:25 issue:1 pages:77-99 ispartof: location:England status: published

To study the dynamic interplay between different component processes involved in the identification of fragmented object outlines, a discrete-identification paradigm was used in which the masked presentation duration of fragmented object outlines was repeatedly increased until correct naming occurred. Survival analysis was used to investigate whether and when different types of information such as contour integration cues (proximity, collinearity, and fragment density), fragment properties (low versus high curvature), stimulus complexity (global symmetry, number and saliency of the parts), and memory factors (natural versus artifactual), influenced the timing of identification. The results showed that the importance of these different types of information can change over the time course of object identification, indicating so-called time-course contingencies. Most importantly, the straight segments of a contour played a larger role for complex outlines with high part saliency during early (bottom-up) grouping processes, whereas the curved segments of object outlines were more important during later (top-down) matching processes for simpler outlines with lower part saliency. This new insight can explain why different studies on shape-based object identification have produced seemingly contradictory results. ispartof: Journal of Experimental Psychology: Human Peception and Performance vol:35 issue:3 pages:661-687 ispartof: location:United States status: published

During visual object categorization, a match must be found between the input image and stored information about basic-level categories. Graf (2002) suggested the involvement of analog transformational, shape-changing processes to align the memory representation of the category with the perceptual representation of the current stimulus. Here we compare the predictions of alignment models with exemplar-based models using morphing between four exemplar outlines within each of eleven categories. Overall, with increasing transformational distance between two exemplars of the same category, reaction times to decide whether they belong to the same category in a sequential matching paradigm increased, while rated similarity between both exemplars decreased. However, in contrast to alignment accounts, exemplar-based accounts (a) can correctly predict the observed dissociation between typicality and categorization time, and allow the observed (b) deviations from sequential additivity, and (c) non-linear relations between transformational distance and rated similarity. By discussing integrations of exemplar-based theories with neglected processes such as information accumulation, response competition, response priming, and gain-modulation, a view of the recognition process from input to response emerges, which increases the validity and scope of modern exemplar-based categorization and recognition models. ispartof: Perception vol:37 issue:12 pages:1822-1849 ispartof: location:United States status: published

status: published

The ease and efficiency with which we perceive objects in daily life masks the complexity of the processes involved. The main goal of my doctoral research was to enhance our understanding of the complex interplay between perceptual organization and object recognition. To this end, we investigated the dynamic interplay between different component processes of object recognition, and their temporal dynamics. In the first part of this thesis, I present three behavioral studies focusing on the role of different features of an objects shape (e.g., local contour information). In the second part of this thesis, I present our neuropsychological work, including a case study and the development of a new screening test for perceptual organization (L-POST). In a first behavioral study (Chapter 2), we used a dynamic build-up paradigm with fragmented outlines to investigate the interplay between grouping and matching processes during object recognition. We tried to disentangle the influence of configural shape information such as symmetry and complexity, and local contour information. Our results revealed a dominance of global configural properties during early grouping processes, compared to local fragment properties (i.e., fragment curvature) or more local configural properties (e.g., proximity). In the following two behavioral studies (Chapters 3 & 4), we investigated the role of within trial top-down processing when people recognize objects, and emphasizing the role of candidate interpretations. We focused on the contribution of different spatial frequencies in their formation, and the time-windows in which they are most effective. In a pre-study with long exposures, participants were asked to list all potential objects that could have given rise to the low-pass filtered pictures. In two priming experiments with short exposures, participants had to classify the objects as natural or manmade, or they had to name them. The results showed that the number of candidate interpretations generated by the low-spatial frequencies of the target influences the occurrence of recognition responses, but only when response latency is long, and all candidates belong to the same category as the target. In Chapter 5, we investigated how global shape information and local contour information influence object identification in a patient with a category-specific impairment for natural objects. We used systematically degraded stimuli (fragmented outlines) to manipulate the quality of the visual representation of the object. The paradigm was similar to the one used in our first behavioral study (Chapter 2). We found evidence that category-specific deficits for natural objects do not only arise at a semantic level, but also reflect poor visual grouping and access to visual knowledge. In Chapter 6, I describe the Leuven Perceptual Organization Screening Test (L-POST) in which a wide range of processes of perceptual organization are measured. The L-POST consists of 15 subtests screening for various aspects of perceptual organization, e.g. figure-ground segregation, contour integration, or local and global processing. To reduce cognitive load, a matching-to-sample task is used for all test items: Participants have to indicate the alternative that is the most similar to the target stimulus. This computerized test can be administered in 30-45 minutes, and a neglect friendly version is also available. To conclude, the research of my doctoral project has enhanced our understanding of the temporal dynamics of different component processes during object recognition. Moreover, our results emphasize the perceptual value of coarse shape information in the activation of candidate interpretations during object recognition. More generally, our results highlight the need to control for the moment of responding in behavioral studies by using an appropriate statistical technique such as survival analysis. nrpages: 196 status: published