Deficits in the Perception of Discontinuities from Motion Cues (clicking on the reference numbers in the text below will bring you to the reference and abstract) Double dissociation of deficits. In 18 patients with unilateral brain lesions we demonstrated a double dissociation of deficits on motion coherence and motion discontinuity detection in random dot kinematograms (RDKs). The stimuli consisted of sparse RDKs as illustrated in panels e) and i) of the Core Motion Tests. This results provides further support to our previously reported findings in two patients which suggested that, contrary to several popular computational theories of motion discontinuity, coherence (which requires global integration) and discontinuity are not computed simultaneously (nor is discontinuity computed at a stage that follows coherence computation). In 4 patients with bilateral lesions [10] and in 12 additional patients with unilateral lesions in the dorsal pathway, we found a similar dissociation when the stimuli embodying the motion discontinuity consisted of dense RDKs. A computational model for motion discontinuity. We (Vaina and Sundreaswaran) developed and implemented a local model that explains the patients’ data and tested the model with the same stimuli that were used in the psychophysical studies of patients’ performance on these tasks. The goal of the local model was to detect motion discontinuity based on measurements within a small aperture. Our model extends Nakayama&Loomis’ model for the situation when only a local projection of the optic flow is available (due to the aperture problem). We used a voting scheme where each normal flow votes a set of directions and the resulting approximation to the convexity function of Nakaayma&Loomis is thresholded to find locations with significant motion discontinuity. The local projection, termed normal flow, can be computed easily from two or more frames of an image sequence. The aperture problem posits that only the component of optic flow along the local intensity gradient direction can be computed. To address this information loss, we extended the Nakayama-Loomis model in the following way. For each direction considered in the Nakayama-Loomis model the projection of optic flow vector in that direction was used in calculating the value of the convexity function. Since our model did not compute optic flow, we devised a voting scheme so that every normal flow vector voted for a set of directions. Without knowing which is the correct optic flow vector, it is impossible to calculate the magnitude of the components. Given this limitation, we treated all components equally hence every component within a 90° range of the normal flow vector gets a vote. To compute the convexity function for each direction di we added the votes within the central region C (NiC) and subtracted k times the total votes within the surrounding region S (NiS). After squaring this difference, we summed the result for all directions (eq. below). The resulting approximation to the convexity function of Nakayama and Loomis was thresholded to find locations with significant motion discontinuities. On a set of dense patterns we showed that the model computes discontinuity, and thus motion discontinuity can be detected by using local motion processes. We suggest that such local computation may be used by the visual system when the global motion mechansims are impaire, or even when the local motion cue is strong and available. Publications 1. Vaina LM, Gryzwacz NM, LeMay M, Bienfang D, Wolpow E “Perception of Motion Discontinuities in Patients with Selective Motion Deficits” in High-level motion processing: computational, neurobiological, and psychophysical perspectives, T. Watanabe (ed), MIT Press, 1998; pp.213-247. back to Research last update: 12/10/98