GMM-IL: Image Classification using Incrementally Learnt, Independent Probabilistic Models for Small Sample Sizes

TL;DR

This paper introduces GMM-IL, a novel image classification approach that uses independent probabilistic models for each class, enabling incremental learning with small samples without catastrophic forgetting.

Contribution

It proposes a two-stage architecture combining visual features with Gaussian Mixture Models for class representation, improving incremental learning and small sample accuracy.

Findings

Outperforms Softmax-based classifiers on small sample sizes

No catastrophic forgetting when learning new classes

Achieves higher accuracy and F1 scores in imbalanced class scenarios

Abstract

Current deep learning classifiers, carry out supervised learning and store class discriminatory information in a set of shared network weights. These weights cannot be easily altered to incrementally learn additional classes, since the classification weights all require retraining to prevent old class information from being lost and also require the previous training data to be present. We present a novel two stage architecture which couples visual feature learning with probabilistic models to represent each class in the form of a Gaussian Mixture Model. By using these independent class representations within our classifier, we outperform a benchmark of an equivalent network with a Softmax head, obtaining increased accuracy for sample sizes smaller than 12 and increased weighted F1 score for 3 imbalanced class profiles in that sample range. When learning new classes our classifier exhibits no catastrophic forgetting issues and only requires the new classes' training images to be present. This enables a database of growing classes over time which can be visually indexed and reasoned over.