My general research interest is to understand how different brain areas process various types of information to construct meaningful object percepts and concepts. Also, I am interested in how object representations are constrained by the interactions between different factors (e.g., visual, semantic, contextual properties). To address these questions, I am currently investigating how several aspects of conceptual knowledge contribute to an appropriate context that improves perceptual speed and accuracy. Such context-dependent effects can illustrate how conceptual knowledge interacts with perception. I also examine the interactions between visual appearance and conceptual knowledge from first impressions to skilled object recognition. I study these interactions using behavioral and neural imaging methods. I am also interested in studying the interactions between the ventral and dorsal visual streams in object processing. One ongoing project is to clarify the roles of these visual areas in object recognition and mental rotation, following up the recent behavioral findings on dissociating the effects of image similarity and angular disparity in these processes (Cheung, Hayward, & Gauthier, 2009). Another aspect of my research examines holistic processing, which is often observed during processing of faces or objects of expertise and reflects an interactive nature between different parts of a stimulus. Please see below for more details about this project.  Nature, Locus, Measures and Interference of Holistic Processing Holistic processing reveals that perception of a part of a stimulus is strongly influenced by the presence of other parts. This research project investigates the nature and locus of such processing and examines the different measures used in the literature. Nature: Holistic processing is often measured by failures of selective attention to a stimulus part while ignoring the other parts in a composite task (Young, Hellawell, & Hay, 1987). This effect may appear quite similar to other selective attention effects which reveal response interference, such as the Stroop effect. One main goal of this project is to dissociate holistic processing from response interference. Using a naming paradigm, we were able to separate holistic processing and response interference and showed that failures of selective attention to part of a face indeed reveal holistic processing instead of response interference (Richler, Cheung, Wong, & Gauthier, 2009; Cheung, Richler, Phillips, Gauthier, 2011). Locus: Although holistic processing is dissociated from response interference, it does not necessarily imply that holistic processing arises from a perceptual locus, where various features may be represented as a whole unit. Instead, features could be represented independently, but perceptual decision to a feature may be influenced by others, resulting in the holistic effects (for a review, see Cheung & Gauthier, 2009). The application of General Recognition Theory may be useful to differentiate the perceptual vs. decisional loci for holistic effects (Richler, Gauthier, Wenger, & Palmeri, 2008), but much work is still needed to resolve this issue. Measures: Different measures of holistic processing used in the literature often produce conflicting results. Thus, we opted to find out the appropriate way to measure holistic processing. Recently, we demonstrated that previous results that only examined hit rates might be susceptible to response bias (Cheung, Richler, Palmeri, & Gauthier, 2008; Richler, Cheung, & Gauthier, in press). Because differential response biases have been found in the composite task for many different manipulations, such as alignment, congruency, inversion and spatial frequency (Cheung & Gauthier, 2009), we suggest that it is important to use an unbiased task (e.g., 2AFC) or measure (e.g., d') and adopt a design that minimizes the confounds (e.g., use the complete design rather than the partial design, see Gauthier & Bukach, 2007; Cheung et al., 2008; Richler, Cheung, & Gauthier, 2011). Competition in Working Memory: Holistic processing suffers from perceptual limitations, such that only very few faces can be processed holistically at a time, and that holistic processing for a target face is susceptible to interference from other faces or objects of expertise (Palermo & Rhodes, 2002; Gauthier et al., 2003). Because working memory capacity is also severely limited (e.g., Luck & Vogel, 1997), we attempted to understand the nature of competition in working memory on holistic face processing. We found that additional faces from the same category as the target face (e.g., adult faces) maintained in working memory reduce holistic processing for a target face, but faces from a different category (e.g., adult vs. child faces) are less likely to compete. But maintenance of non-face objects in working memory do not compete with holistic face processing, regardless of the level of expertise one has with the object category. Thus, interference in working memory on holistic face processing is likely due to the perceived similarity among the target faces and the items stored in working memory. We suggest that the nature of competition on holistic processing in working memory is different from that at the perceptual level (Cheung & Gauthier, 2010). |
|
|
Research |