Motion contrast is important for object detection, yet sensitivity to its underlying spatiotemporal properties remains largely unexplored. Speed tuning may differ for figures defined by direction versus speed contrast, due to the distinct ecological context associated with each figure type (i.e., a moving object and motion parallax, respectively). However, similar response tuning for motion contrast magnitude has been observed across different types of motion-defined figures (Fesi et al., 2011), indicating that their cortical mechanisms may exhibit cue-invariance. In this study, we compared steady-state visual evoked potential (SSVEP) responses of n=20 adults (10 female, mean age: 19.5 yrs) to time-varying dot displays in which four figure regions emerged from and disappeared into background dots at a specified rate (1.2 Hz, F1). Figures defined by direction or global dot coherence contrast moved at 2, 4, and 16 degrees/s. Figures defined by speed contrast had background dots moving at 2, 4, and 16 deg/s, but featured a consistently faster figure speed relative to background. For the direction and coherence conditions, amplitudes at 1F1 peaked at 16 deg/s, while speed-defined figures elicited peak activation at 4 deg/s. Topographic maps of activation reveal a uniform medial occipital distribution of the 1F1 response for nearly all conditions, except for speed-defined figures. Here, the distribution shifted from medial occipital to centro-frontal channels as figure and ground increased in speed. Our results indicate that: 1) the processing of direction and coherence-defined figures are largely similar in terms of tuning and distribution along the scalp, and 2) responses for speed contrast are similar to those for other types of motion contrast, but only when figure and background move at slower speeds. This discrepancy may reflect sensitivity to the natural statistics of objects in central visual field, but also serves to qualify interpretations of cue-invariant processing of motion-defined figures.
Meeting abstract presented at VSS 2012