Conditional Variational Capsule Network for Open Set Recognition

TL;DR

This paper introduces a novel conditional variational capsule network that enhances open set recognition by encouraging compact class-specific feature distributions, achieving state-of-the-art results in detecting unknown classes.

Contribution

It proposes a new capsule network framework using variational autoencoders with Gaussian priors to improve open set recognition capabilities.

Findings

Achieved state-of-the-art results on multiple datasets.

Effectively detects unknown classes with high accuracy.

Controls feature compactness to distinguish known and unknown samples.

Abstract

In open set recognition, a classifier has to detect unknown classes that are not known at training time. In order to recognize new categories, the classifier has to project the input samples of known classes in very compact and separated regions of the features space for discriminating samples of unknown classes. Recently proposed Capsule Networks have shown to outperform alternatives in many fields, particularly in image recognition, however they have not been fully applied yet to open-set recognition. In capsule networks, scalar neurons are replaced by capsule vectors or matrices, whose entries represent different properties of objects. In our proposal, during training, capsules features of the same known class are encouraged to match a pre-defined gaussian, one for each class. To this end, we use the variational autoencoder framework, with a set of gaussian priors as the approximation for the posterior distribution. In this way, we are able to control the compactness of the features of the same class around the center of the gaussians, thus controlling the ability of the classifier in detecting samples from unknown classes. We conducted several experiments and ablation of our model, obtaining state of the art results on different datasets in the open set recognition and unknown detection tasks.