DetailVerifyBench: A Benchmark for Dense Hallucination Localization in Long Image Captions

TL;DR

DetailVerifyBench is a new, challenging benchmark with dense annotations for evaluating hallucination localization in long image captions, covering diverse domains and extensive contexts.

Contribution

It introduces the first comprehensive benchmark with token-level hallucination annotations for long image captioning across multiple domains.

Findings

Benchmark contains 1,000 images with over 200-word captions.

Dense, token-level annotations of hallucination types are provided.

Available at https://zyx-hhnkh.github.io/DetailVerifyBench/.

Abstract

Accurately detecting and localizing hallucinations is a critical task for ensuring high reliability of image captions. In the era of Multimodal Large Language Models (MLLMs), captions have evolved from brief sentences into comprehensive narratives, often spanning hundreds of words. This shift exponentially increases the challenge: models must now pinpoint specific erroneous spans or words within extensive contexts, rather than merely flag response-level inconsistencies. However, existing benchmarks lack the fine granularity and domain diversity required to evaluate this capability. To bridge this gap, we introduce DetailVerifyBench, a rigorous benchmark comprising 1,000 high-quality images across five distinct domains. With an average caption length of over 200 words and dense, token-level annotations of multiple hallucination types, it stands as the most challenging benchmark for precise hallucination localization in the field of long image captioning to date. Our benchmark is available at https://zyx-hhnkh.github.io/DetailVerifyBench/.