This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. Primary support for the subproject and the subproject's principal investigator may have been provided by other sources, including other NIH sources. The Total Cost listed for the subproject likely represents the estimated amount of Center infrastructure utilized by the subproject, not direct funding provided by the NCRR grant to the subproject or subproject staff. Moving faces and bodies provide a wealth of information that can be used to facilitate recognition. How then is motion used for recognition (including the social inferences that are the focus of the current proposal)? O'Toole et al. discuss three hypotheses which offer mutually complementary perspectives on how facial motion in particular is processed by the human visual system: (1) Motion may provide supplemental information for recognition, (2) Motion may facilitate the recovery of 3D shape ,andshape, and (3) Motion may provide social information. The first two hypotheses apply to moving object recognition in general, while all three are applicable to both moving faces and bodies. Each hypothesis suggests a specific role for how motion functions in the service of recognition and all are relevant for the """"""""thin-slices"""""""" problem domain. Recent results indicate the extent to which all three aspects of face and body movement may affect individuation, suggesting potentially useful methodologies for the examination of thin-slice perception. In general, previous results identifying aspects of motion that facilitate face and body recognition and manipulations of spatiotemporal properties that interfere with it provide a foundation for identifying """"""""critical features"""""""" of thin-slice stimuli. Familiar face recognition does appear to benefit from motion information in many cases, while unfamiliar face recognition less consistently improves with the addition of motion. Motion advantages are particularly evident when image quality is degraded either through negation or inversion. This could be taken to mean that the additional information provided by facial motion is redundant with high-quality static appearance, but control experiments comparing the recognition of moving faces with static """"""""slideshows"""""""" of multiple images suggest that it is not merely enough to see all of the images in a sequence. This is further established by results from point-light faces that offer very little static information, yet support multiple recognition tasks. Motion thus appears to impact face processing beyond the sum of the individual static images it offers. What then, is the nature of the contribution? In terms of behavioral performance, some of the most interesting results concern how manipulations of the temporal characteristics of dynamic faces impact performance. Specifically, the reversal of face motion is known to disrupt performance, as is modulation of the playback rate. The former result indicates that facial gestures are encoded in a directed sequence (as are rigidly rotating objects while the latter suggests that there may be """"""""tuning"""""""" for movements at natural speeds. Also, Roark et al. have provided compelling evidence that structure-from-motion provides little in the way of additionally useful information for face recognition, underscoring the need to describe the processing of facial gestures without assuming that the recovery of 3D shape is a key goal. Recently, representational strategies for describing facial gestures have been explored by computer vision researchers seeking to build automated systems for recognizing idiosyncratic movements. Early results in this effort are promising, and do not rely upon 3D shape recovery, further weakening the case for structure-from-motion as a fundamental part of how motion facilitates the recognition of moving faces. Dr. Alice O'Toole serves as Dr. Balas'Individual Project Mentor. Dr. O'Toole will visit Dr. Balas'laboatory once per year (and vice versa), and will assist him in the design and interpretation of experiments, and with the preparation and critical review of grant proposals.
Showing the most recent 10 out of 84 publications
