Kernelized Multiview Projection

TL;DR

Kernelized Multiview Projection (KMP) is an unsupervised spectral embedding method that fuses multiple feature views into a low-dimensional, discriminative space, effectively addressing the out-of-sample problem in multiview data analysis.

Contribution

The paper introduces KMP, a novel kernelized spectral embedding algorithm that efficiently combines multiple feature views into a meaningful low-dimensional space while solving out-of-sample issues.

Findings

KMP effectively fuses multiview features into a discriminative embedding.

KMP addresses the out-of-sample problem in multiview spectral embedding.

Experiments show KMP's superior performance on image datasets.

Abstract

Conventional vision algorithms adopt a single type of feature or a simple concatenation of multiple features, which is always represented in a high-dimensional space. In this paper, we propose a novel unsupervised spectral embedding algorithm called Kernelized Multiview Projection (KMP) to better fuse and embed different feature representations. Computing the kernel matrices from different features/views, KMP can encode them with the corresponding weights to achieve a low-dimensional and semantically meaningful subspace where the distribution of each view is sufficiently smooth and discriminative. More crucially, KMP is linear for the reproducing kernel Hilbert space (RKHS) and solves the out-of-sample problem, which allows it to be competent for various practical applications. Extensive experiments on three popular image datasets demonstrate the effectiveness of our multiview embedding algorithm.