# Multimodal Subspace Support Vector Data Description

## TL;DR

This paper introduces a new multimodal subspace method for one-class classification that optimally combines data from multiple modalities into a shared subspace, improving data description performance.

## Contribution

It presents a novel iterative transformation approach with regularization strategies for multimodal data projection, applicable in both linear and nonlinear forms.

## Key findings

- Outperforms competing methods on four out of five datasets.
- Effective in single and multimodal data fusion scenarios.
- Provides both linear and nonlinear formulations.

## Abstract

In this paper, we propose a novel method for projecting data from multiple modalities to a new subspace optimized for one-class classification. The proposed method iteratively transforms the data from the original feature space of each modality to a new common feature space along with finding a joint compact description of data coming from all the modalities. For data in each modality, we define a separate transformation to map the data from the corresponding feature space to the new optimized subspace by exploiting the available information from the class of interest only. We also propose different regularization strategies for the proposed method and provide both linear and non-linear formulations. The proposed Multimodal Subspace Support Vector Data Description outperforms all the competing methods using data from a single modality or fusing data from all modalities in four out of five datasets.