Multi-Object Hallucination in Vision-Language Models

TL;DR

This paper investigates how large vision-language models hallucinate multiple objects in images, introducing a new evaluation protocol and analyzing factors influencing these hallucinations to improve model reliability.

Contribution

The work systematically studies multi-object hallucination in LVLMs, introduces the ROPE evaluation protocol, and provides comprehensive analysis of factors affecting hallucination behaviors.

Findings

LVLMs hallucinate more with multiple objects than single objects

Object class distribution influences hallucination patterns

Data factors, salience, and model behavior affect hallucination likelihood

Abstract

Large vision language models (LVLMs) often suffer from object hallucination, producing objects not present in the given images. While current benchmarks for object hallucination primarily concentrate on the presence of a single object class rather than individual entities, this work systematically investigates multi-object hallucination, examining how models misperceive (e.g., invent nonexistent objects or become distracted) when tasked with focusing on multiple objects simultaneously. We introduce Recognition-based Object Probing Evaluation (ROPE), an automated evaluation protocol that considers the distribution of object classes within a single image during testing and uses visual referring prompts to eliminate ambiguity. With comprehensive empirical studies and analysis of potential factors leading to multi-object hallucination, we found that (1). LVLMs suffer more hallucinations when focusing on multiple objects compared to a single object. (2). The tested object class distribution affects hallucination behaviors, indicating that LVLMs may follow shortcuts and spurious correlations. (3). Hallucinatory behaviors are influenced by data-specific factors, salience and frequency, and model intrinsic behaviors. We hope to enable LVLMs to recognize and reason about multiple objects that often occur in realistic visual scenes, provide insights, and quantify our progress towards mitigating the issues.