MMF: A loss extension for feature learning in open set recognition
This work addresses the problem of identifying unknown classes in applications like malware detection, offering an incremental improvement to existing loss functions.
The paper tackles open set recognition by proposing a loss extension for neural networks to create polar representations for known classes, improving separability from unknown classes. The method significantly enhances performance across two loss functions and datasets, with one loss function showing superior training time and accuracy.
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.