MMF: A loss extension for feature learning in open set recognition

TL;DR

This paper introduces MMF, a loss extension for neural networks that enhances feature separation in open set recognition, improving accuracy and training efficiency across different loss functions and datasets.

Contribution

The paper proposes a novel loss extension called MMF that can be integrated into various loss functions to improve feature discrimination in open set recognition.

Findings

Significantly improves performance of different loss functions on diverse datasets.

Enhances the discriminability of class representations for better open set recognition.

One loss function with MMF outperforms others in training time and accuracy.

Abstract

Open set recognition (OSR) is the problem of classifying the known classes, meanwhile identifying the unknown classes when the collected samples cannot exhaust all the classes. There are many applications for the OSR problem. For instance, the frequently emerged new malware classes require a system that can classify the known classes and identify the unknown malware classes. In this paper, we propose an add-on extension for loss functions in neural networks to address the OSR problem. Our loss extension leverages the neural network to find polar representations for the known classes so that the representations of the known and the unknown classes become more effectively separable. Our contributions include: First, we introduce an extension that can be incorporated into different loss functions to find more discriminative representations. Second, we show that the proposed extension can significantly improve the performances of two different types of loss functions on datasets from two different domains. Third, we show that with the proposed extension, one loss function outperforms the others in terms of training time and model accuracy.