Stacked Capsule Autoencoders

06/17/2019
by   Adam R. Kosiorek, et al.
4

An object can be seen as a geometrically organized set of interrelated parts. A system that makes explicit use of these geometric relationships to recognize objects should be naturally robust to changes in viewpoint, because the intrinsic geometric relationships are viewpoint-invariant. We describe an unsupervised version of capsule networks, in which a neural encoder, which looks at all of the parts, is used to infer the presence and poses of object capsules. The encoder is trained by backpropagating through a decoder, which predicts the pose of each already discovered part using a mixture of pose predictions. The parts are discovered directly from an image, in a similar manner, by using a neural encoder, which infers parts and their affine transformations. The corresponding decoder models each image pixel as a mixture of predictions made by affine-transformed parts. We learn object- and their part-capsules on unlabeled data, and then cluster the vectors of presences of object capsules. When told the names of these clusters, we achieve state-of-the-art results for unsupervised classification on SVHN (55 state-of-the-art on MNIST (98.5

READ FULL TEXT

page 4

page 13

research
12/06/2019

Geometric Capsule Autoencoders for 3D Point Clouds

We propose a method to learn object representations from 3D point clouds...
research
04/07/2020

Capsule Networks – A Probabilistic Perspective

'Capsule' models try to explicitly represent the poses of objects, enfor...
research
04/30/2021

DPR-CAE: Capsule Autoencoder with Dynamic Part Representation for Image Parsing

Parsing an image into a hierarchy of objects, parts, and relations is im...
research
03/11/2021

Inference for Generative Capsule Models

Capsule networks (see e.g. Hinton et al., 2018) aim to encode knowledge ...
research
11/27/2020

Unsupervised part representation by Flow Capsules

Capsule networks are designed to parse an image into a hierarchy of obje...
research
09/07/2022

Inference and Learning for Generative Capsule Models

Capsule networks (see e.g. Hinton et al., 2018) aim to encode knowledge ...
research
11/29/2022

Testing GLOM's ability to infer wholes from ambiguous parts

The GLOM architecture proposed by Hinton [2021] is a recurrent neural ne...

Please sign up or login with your details

Forgot password? Click here to reset