Timezone: »

 
Spotlight
Unsupervised Part Representation by Flow Capsules
Sara Sabour Rouh Aghdam · Andrea Tagliasacchi · Soroosh Yazdani · Geoffrey Hinton · David Fleet

Thu Jul 22 06:30 PM -- 06:35 PM (PDT) @
in Representation Learning 5 »

Capsule networks aim to parse images into a hierarchy of objects, parts and relations. While promising, they remain limited by an inability to learn effective low level part descriptions. To address this issue we propose a way to learn primary capsule encoders that detect atomic parts from a single image. During training we exploit motion as a powerful perceptual cue for part definition, with an expressive decoder for part generation within a layered image model with occlusion. Experiments demonstrate robust part discovery in the presence of multiple objects, cluttered backgrounds, and occlusion. The learned part decoder is shown to infer the underlying shape masks, effectively filling in occluded regions of the detected shapes. We evaluate FlowCapsules on unsupervised part segmentation and unsupervised image classification.

Keywords: Embedding and Representation learning

Author Information

Sara Sabour Rouh Aghdam (Google)
Andrea Tagliasacchi (Google Inc.)
Soroosh Yazdani (Google Inc.)
Geoffrey Hinton (Google)
David Fleet (University of Toronto)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors