Distance Based Image Classification: A solution to generative classification's conundrum?

TL;DR

This paper introduces a novel distance-based generative classification method that emphasizes semantic features over non-semantic noise, achieving high accuracy and scalability compared to traditional discriminative classifiers.

Contribution

It proposes a new generative model incorporating hierarchical semantic factors and noise, leading to a modified nearest-neighbor classifier called distance classification.

Findings

Achieves high accuracy as a generative classifier

Scales well with the number of classes

Supports incremental updates

Abstract

Most classifiers rely on discriminative boundaries that separate instances of each class from everything else. We argue that discriminative boundaries are counter-intuitive as they define semantics by what-they-are-not; and should be replaced by generative classifiers which define semantics by what-they-are. Unfortunately, generative classifiers are significantly less accurate. This may be caused by the tendency of generative models to focus on easy to model semantic generative factors and ignore non-semantic factors that are important but difficult to model. We propose a new generative model in which semantic factors are accommodated by shell theory's hierarchical generative process and non-semantic factors by an instance specific noise term. We use the model to develop a classification scheme which suppresses the impact of noise while preserving semantic cues. The result is a surprisingly accurate generative classifier, that takes the form of a modified nearest-neighbor algorithm; we term it distance classification. Unlike discriminative classifiers, a distance classifier: defines semantics by what-they-are; is amenable to incremental updates; and scales well with the number of classes.