Semantically Distributed Robust Optimization for Vision-and-Language Inference

TL;DR

This paper introduces SDRO, a model-agnostic approach that enhances vision-and-language models' robustness to linguistic variations through distributed robust optimization and ensembling, improving performance on multiple benchmarks.

Contribution

The paper proposes SDRO, a novel method integrating linguistic transformations into training and inference for increased robustness in vision-and-language tasks.

Findings

Improves robustness to linguistic adversarial attacks

Enhances performance on NLVR$^2$ and VIOLIN datasets

Shows generalizability to other V extbar L tasks

Abstract

Analysis of vision-and-language models has revealed their brittleness under linguistic phenomena such as paraphrasing, negation, textual entailment, and word substitutions with synonyms or antonyms. While data augmentation techniques have been designed to mitigate against these failure modes, methods that can integrate this knowledge into the training pipeline remain under-explored. In this paper, we present \textbf{SDRO}, a model-agnostic method that utilizes a set linguistic transformations in a distributed robust optimization setting, along with an ensembling technique to leverage these transformations during inference. Experiments on benchmark datasets with images (NLVR$^2$) and video (VIOLIN) demonstrate performance improvements as well as robustness to adversarial attacks. Experiments on binary VQA explore the generalizability of this method to other V\&L tasks.