Multi-View Learning in the Presence of View Disagreement

TL;DR

This paper introduces a multi-view learning method that detects and filters out view disagreement using a conditional entropy criterion, improving learning performance in noisy or corrupted data scenarios.

Contribution

It proposes a novel approach for identifying and removing view disagreement, enhancing the robustness of multi-view learning methods.

Findings

Detection of view disagreement improves learning accuracy.

Filtering out disagreeing views enhances performance on synthetic and real datasets.

Method effectively handles noise, occlusion, and view corruption.

Abstract

Traditional multi-view learning approaches suffer in the presence of view disagreement,i.e., when samples in each view do not belong to the same class due to view corruption, occlusion or other noise processes. In this paper we present a multi-view learning approach that uses a conditional entropy criterion to detect view disagreement. Once detected, samples with view disagreement are filtered and standard multi-view learning methods can be successfully applied to the remaining samples. Experimental evaluation on synthetic and audio-visual databases demonstrates that the detection and filtering of view disagreement considerably increases the performance of traditional multi-view learning approaches.