University of Cambridge > Talks.cam > Craik Club > Perceptual Organization of Shape

Perceptual Organization of Shape

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact John Mollon.

Humans are very good at rapidly detecting salient natural shapes such as animals in complex scenes. This suggests that our visual system is finely tuned to the statistics of natural shape, and this is borne out by recent evidence for efficient coding of local shape information. In computational models the prevailing view has been that these local regularities drive the perceptual organization of global shape in feed-forward fashion. However, feed-forward computer vision algorithms fall well below human levels of performance, and recent results suggest that perceptual organization is also drive by global cues. This suggests a recurrent architecture in which higher areas of the object pathway generate shape hypotheses that condition grouping processes in early visual areas. As a candidate generative model for shape coding, I will explore a framework in which shape is progressively represented as a sequence of deformations called formlets. I will compare this formlet representation with alternative theories in terms of accuracy in completing fragmented shape hypotheses that may be generated by the early visual system, and in terms of capturing the shape information required for recognition.

This talk is part of the Craik Club series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2020 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity