Robustness in Multimodal Learning under Train-Test Modality Mismatch
Brandon McKinzie, Vaishaal Shankar, Joseph Yitan Cheng, Yinfei Yang, Jonathon Shlens, Alexander T Toshev
Abstract
Multimodal learning is defined as learning over multiple heterogeneous input modalities such as video, audio, and text. In this work, we are concerned with understanding how models behave as the type of modalities differ between training and deployment, a situation that naturally arises in many applications of multimodal learning to hardware platforms. We present a multimodal robustness framework to provide a systematic analysis of common multimodal representation learning methods. Further, we identify robustness short-comings of these approaches and propose two intervention techniques leading to $1.5\times$-$4\times$ robustness improvements on three datasets, AudioSet, Kinetics-400 and ImageNet-Captions. Finally, we demonstrate that these interventions better utilize additional modalities, if present, to achieve competitive results of $44.2$ mAP on AudioSet 20K.
BibTeX
@inproceedings{icml2023_robustnessinmult,
title = {Robustness in Multimodal Learning under Train-Test Modality Mismatch},
author = {Brandon McKinzie and Vaishaal Shankar and Joseph Yitan Cheng and Yinfei Yang and Jonathon Shlens and Alexander T Toshev},
booktitle = {ICML 2023},
year = {2023}
}