VL-Uncertainty: Detecting Hallucination in Large Vision-Language Model via Uncertainty Estimation
This addresses safety concerns in LVLMs by reducing the need for human annotation in hallucination detection, though it is an incremental advancement in uncertainty estimation methods.
The paper tackles the problem of detecting hallucinations in large vision-language models (LVLMs) by introducing VL-Uncertainty, an uncertainty-based framework that uses prediction variance across perturbed prompts and cluster distribution entropy, achieving significant performance improvements over baselines in experiments on 10 LVLMs across four benchmarks.
Given the higher information load processed by large vision-language models (LVLMs) compared to single-modal LLMs, detecting LVLM hallucinations requires more human and time expense, and thus rise a wider safety concerns. In this paper, we introduce VL-Uncertainty, the first uncertainty-based framework for detecting hallucinations in LVLMs. Different from most existing methods that require ground-truth or pseudo annotations, VL-Uncertainty utilizes uncertainty as an intrinsic metric. We measure uncertainty by analyzing the prediction variance across semantically equivalent but perturbed prompts, including visual and textual data. When LVLMs are highly confident, they provide consistent responses to semantically equivalent queries. However, when uncertain, the responses of the target LVLM become more random. Considering semantically similar answers with different wordings, we cluster LVLM responses based on their semantic content and then calculate the cluster distribution entropy as the uncertainty measure to detect hallucination. Our extensive experiments on 10 LVLMs across four benchmarks, covering both free-form and multi-choice tasks, show that VL-Uncertainty significantly outperforms strong baseline methods in hallucination detection.