Mitigating Object Hallucination in MLLMs via Data-augmented Phrase-level Alignment

TL;DR

This paper introduces Data-augmented Phrase-level Alignment (DPA), a novel training method that significantly reduces object hallucinations in Multimodal Large Language Models while maintaining their general capabilities.

Contribution

The work proposes DPA, a new loss function that leverages generated phrase-level data augmentation to mitigate hallucinations in MLLMs without sacrificing overall performance.

Findings

DPA reduces hallucination rates by up to 4.2% on image description tasks.

Finetuned MLLMs with DPA, called HALVA, improve F1 scores by up to 13.4% on hallucination visual question-answering.

DPA preserves the general vision-language capabilities of MLLMs.

Abstract

Despite their significant advancements, Multimodal Large Language Models (MLLMs) often generate factually inaccurate information, referred to as hallucination. In this work, we address object hallucinations in MLLMs, where information is generated about an object not present in the input image. We introduce Data-augmented Phrase-level Alignment (DPA), a novel loss which can be applied to instruction-tuned off-the-shelf MLLMs to mitigate hallucinations, while preserving their general vision-language capabilities. To fine-tune MLLMs with DPA, we first generate a set of `hallucinated' and `correct' response pairs through generative data augmentation by selectively altering the ground-truth information of the correct responses at a phrase level. The DPA loss is then used to train MLLMs to reduce the likelihood of hallucinated phrases compared to the correct ones. Our thorough evaluation on various benchmarks confirms the effectiveness of DPA in mitigating hallucination while retaining the out-of-the-box performance of the MLLMs on general tasks. For instance, MLLMs finetuned with DPA, which we refer to as Hallucination Attenuated Language and Vision Assistant (HALVA), improve F1 by up to 13.4% on hallucination visual question-answering and reduce the hallucination rate by up to 4.2% on image description tasks.