Reefknot: A Comprehensive Benchmark for Relation Hallucination Evaluation, Analysis and Mitigation in Multimodal Large Language Models

TL;DR

Reefknot is a new benchmark with over 20,000 samples designed to evaluate and reduce relation hallucinations in multimodal large language models, highlighting current limitations and proposing mitigation strategies.

Contribution

The paper introduces Reefknot, a comprehensive benchmark for relation hallucination evaluation and a confidence-based mitigation method for MLLMs.

Findings

Current MLLMs struggle with relation hallucinations.

The proposed mitigation reduces hallucinations by an average of 9.75%.

Reefknot provides a detailed evaluation framework for relation hallucinations.

Abstract

Hallucination issues continue to affect multimodal large language models (MLLMs), with existing research mainly addressing object-level or attribute-level hallucinations, neglecting the more complex relation hallucinations that require advanced reasoning. Current benchmarks for relation hallucinations lack detailed evaluation and effective mitigation, and their datasets often suffer from biases due to systematic annotation processes. To address these challenges, we introduce Reefknot, a comprehensive benchmark targeting relation hallucinations, comprising over 20,000 real-world samples. We provide a systematic definition of relation hallucinations, integrating perceptive and cognitive perspectives, and construct a relation-based corpus using the Visual Genome scene graph dataset. Our comparative evaluation reveals significant limitations in current MLLMs' ability to handle relation hallucinations. Additionally, we propose a novel confidence-based mitigation strategy, which reduces the hallucination rate by an average of 9.75% across three datasets, including Reefknot. Our work offers valuable insights for achieving trustworthy multimodal intelligence.