Variational Distillation for Multi-View Learning

TL;DR

This paper introduces a scalable variational distillation method for multi-view learning that effectively captures shared information, producing compact, predictive representations while handling heterogeneous data without explicit mutual information estimation.

Contribution

It proposes a novel Multi-View Variational Distillation (MV²D) strategy that offers a theoretical, flexible, and scalable solution for multi-view representation learning based on information bottleneck principles.

Findings

Outperforms state-of-the-art methods on multiple benchmarks.

Effectively neutralizes heterogeneity sensitivity in multi-view data.

Produces compact, predictive representations with theoretical guarantees.

Abstract

Information Bottleneck (IB) based multi-view learning provides an information theoretic principle for seeking shared information contained in heterogeneous data descriptions. However, its great success is generally attributed to estimate the multivariate mutual information which is intractable when the network becomes complicated. Moreover, the representation learning tradeoff, {\it i.e.}, prediction-compression and sufficiency-consistency tradeoff, makes the IB hard to satisfy both requirements simultaneously. In this paper, we design several variational information bottlenecks to exploit two key characteristics ({\it i.e.}, sufficiency and consistency) for multi-view representation learning. Specifically, we propose a Multi-View Variational Distillation (MV$^2$D) strategy to provide a scalable, flexible and analytical solution to fitting MI by giving arbitrary input of viewpoints but without explicitly estimating it. Under rigorously theoretical guarantee, our approach enables IB to grasp the intrinsic correlation between observations and semantic labels, producing predictive and compact representations naturally. Also, our information-theoretic constraint can effectively neutralize the sensitivity to heterogeneous data by eliminating both task-irrelevant and view-specific information, preventing both tradeoffs in multiple view cases. To verify our theoretically grounded strategies, we apply our approaches to various benchmarks under three different applications. Extensive experiments to quantitatively and qualitatively demonstrate the effectiveness of our approach against state-of-the-art methods.