Hyperspherical Prototype Networks

01/29/2019
by   Pascal Mettes, et al.
12

This paper introduces hyperspherical prototype networks, which unify regression and classification by prototypes on hyperspherical output spaces. Rather than defining prototypes as the mean output vector over training examples per class, we propose hyperspheres as output spaces to define class prototypes a priori with large margin separation. By doing so, we do not require any prototype updating, we can handle any training size, and the output dimensionality is no longer constrained to the number of classes. Furthermore, hyperspherical prototype networks generalize to regression, by optimizing outputs as an interpolation between two prototypes on the hypersphere. Since both tasks are now defined by the same loss function, they can be jointly optimized for multi-task problems. Experimental evaluation shows the benefits of hyperspherical prototype networks for classification, regression, and their combination.

READ FULL TEXT
research
04/11/2022

ProtoTEx: Explaining Model Decisions with Prototype Tensors

We present ProtoTEx, a novel white-box NLP classification architecture b...
research
10/15/2020

A Theory of Hyperbolic Prototype Learning

We introduce Hyperbolic Prototype Learning, a type of supervised learnin...
research
11/29/2017

A Semantic Loss Function for Deep Learning with Symbolic Knowledge

This paper develops a novel methodology for using symbolic knowledge in ...
research
02/26/2022

Semantic Supervision: Enabling Generalization over Output Spaces

In this paper, we propose Semantic Supervision (SemSup) - a unified para...
research
06/26/2023

ProtoDiff: Learning to Learn Prototypical Networks by Task-Guided Diffusion

Prototype-based meta-learning has emerged as a powerful technique for ad...
research
07/03/2021

Cluster Representatives Selection in Non-Metric Spaces for Nearest Prototype Classification

The nearest prototype classification is a less computationally intensive...
research
04/17/2017

Fast multi-output relevance vector regression

This paper aims to decrease the time complexity of multi-output relevanc...

Please sign up or login with your details

Forgot password? Click here to reset