Beyond Unimodal Learning: The Importance of Integrating Multiple Modalities for Lifelong Learning

TL;DR

This paper emphasizes the importance of integrating multiple modalities in deep neural networks to improve lifelong learning, reduce catastrophic forgetting, and enhance robustness, inspired by human multimodal learning.

Contribution

It introduces a new benchmark for multimodal continual learning, analyzes the role of multiple modalities, and proposes a method for aligning information across modalities.

Findings

Multimodal learning improves robustness and reduces forgetting.

Different modalities have varying resilience to distribution shifts.

The proposed method effectively aligns multimodal data for better inference.

Abstract

While humans excel at continual learning (CL), deep neural networks (DNNs) exhibit catastrophic forgetting. A salient feature of the brain that allows effective CL is that it utilizes multiple modalities for learning and inference, which is underexplored in DNNs. Therefore, we study the role and interactions of multiple modalities in mitigating forgetting and introduce a benchmark for multimodal continual learning. Our findings demonstrate that leveraging multiple views and complementary information from multiple modalities enables the model to learn more accurate and robust representations. This makes the model less vulnerable to modality-specific regularities and considerably mitigates forgetting. Furthermore, we observe that individual modalities exhibit varying degrees of robustness to distribution shift. Finally, we propose a method for integrating and aligning the information from different modalities by utilizing the relational structural similarities between the data points in each modality. Our method sets a strong baseline that enables both single- and multimodal inference. Our study provides a promising case for further exploring the role of multiple modalities in enabling CL and provides a standard benchmark for future research.