Exploring Hallucination of Large Multimodal Models in Video Understanding: Benchmark, Analysis and Mitigation

• URL: http://arxiv.org/abs/2503.19622v1
• Date: Tue, 25 Mar 2025 13:12:17 GMT
• Title: Exploring Hallucination of Large Multimodal Models in Video Understanding: Benchmark, Analysis and Mitigation
• Authors: Hongcheng Gao, Jiashu Qu, Jingyi Tang, Baolong Bi, Yue Liu, Hongyu Chen, Li Liang, Li Su, Qingming Huang,
• Abstract summary: hallucination of large multimodal models (LMMs) provides responses that appear correct but are actually incorrect.<n>This paper aims to study the hallucination problem of LMMs in video modality, which is dynamic and more challenging compared to static modalities like images and text.
• Score: 49.885797244626694
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The hallucination of large multimodal models (LMMs), providing responses that appear correct but are actually incorrect, limits their reliability and applicability. This paper aims to study the hallucination problem of LMMs in video modality, which is dynamic and more challenging compared to static modalities like images and text. From this motivation, we first present a comprehensive benchmark termed HAVEN for evaluating hallucinations of LMMs in video understanding tasks. It is built upon three dimensions, i.e., hallucination causes, hallucination aspects, and question formats, resulting in 6K questions. Then, we quantitatively study 7 influential factors on hallucinations, e.g., duration time of videos, model sizes, and model reasoning, via experiments of 16 LMMs on the presented benchmark. In addition, inspired by recent thinking models like OpenAI o1, we propose a video-thinking model to mitigate the hallucinations of LMMs via supervised reasoning fine-tuning (SRFT) and direct preference optimization (TDPO)-- where SRFT enhances reasoning capabilities while TDPO reduces hallucinations in the thinking process. Extensive experiments and analyses demonstrate the effectiveness. Remarkably, it improves the baseline by 7.65% in accuracy on hallucination evaluation and reduces the bias score by 4.5%. The code and data are public at https://github.com/Hongcheng-Gao/HAVEN.

Related papers

• A Survey of Hallucination in Large Visual Language Models [48.794850395309076]
  The existence of hallucinations has limited the potential and practical effectiveness of LVLM in various fields. The structure of LVLMs and main causes of hallucination generation are introduced. The available hallucination evaluation benchmarks for LVLMs are presented.
  arXiv  Detail & Related papers  (2024-10-20T10:58:58Z)
• From Single to Multi: How LLMs Hallucinate in Multi-Document Summarization [6.37435726278524]
  We investigate how hallucinations manifest in large language models (LLMs) when summarizing topic-specific information from multiple documents. On average, up to 75% of the content in LLM-generated summary is hallucinated, with hallucinations more likely to occur towards the end of the summaries. To understand the characteristics of these hallucinations, we manually evaluate 700+ insights and find that most errors stem from either failing to follow instructions or producing overly generic insights.
  arXiv  Detail & Related papers  (2024-10-17T18:38:53Z)
• FaithBench: A Diverse Hallucination Benchmark for Summarization by Modern LLMs [2.871226288151562]
  This paper introduces FaithBench, a summarization hallucination benchmark comprising challenging hallucinations made by 10 modern LLMs. Our results show GPT-4o and GPT-3.5-Turbo produce the least hallucinations. Even the best hallucination detection models have near 50% accuracies on FaithBench, indicating lots of room for future improvement.
  arXiv  Detail & Related papers  (2024-10-17T04:30:46Z)
• FG-PRM: Fine-grained Hallucination Detection and Mitigation in Language Model Mathematical Reasoning [10.709365940160685]
  Existing approaches primarily detect the presence of hallucinations but lack a nuanced understanding of their types and manifestations. We introduce a comprehensive taxonomy that categorizes the common hallucinations in mathematical reasoning task into six types. We then propose FG-PRM, an augmented model designed to detect and mitigate hallucinations in a fine-grained, step-level manner.
  arXiv  Detail & Related papers  (2024-10-08T19:25:26Z)
• Reefknot: A Comprehensive Benchmark for Relation Hallucination Evaluation, Analysis and Mitigation in Multimodal Large Language Models [13.48296910438554]
  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. We propose a novel confidence-based mitigation strategy, which reduces the hallucination rate by an average of 9.75% across three datasets, including Reefknot.
  arXiv  Detail & Related papers  (2024-08-18T10:07:02Z)
• Hallu-PI: Evaluating Hallucination in Multi-modal Large Language Models within Perturbed Inputs [54.50483041708911]
  Hallu-PI is the first benchmark designed to evaluate hallucination in MLLMs within Perturbed Inputs. Hallu-PI consists of seven perturbed scenarios, containing 1,260 perturbed images from 11 object types. Our research reveals a severe bias in MLLMs' ability to handle different types of hallucinations.
  arXiv  Detail & Related papers  (2024-08-02T16:07:15Z)
• ANAH-v2: Scaling Analytical Hallucination Annotation of Large Language Models [65.12177400764506]
  Large language models (LLMs) exhibit hallucinations in long-form question-answering tasks across various domains and wide applications.<n>Current hallucination detection and mitigation datasets are limited in domains and sizes.<n>This paper introduces an iterative self-training framework that simultaneously and progressively scales up the hallucination annotation dataset.
  arXiv  Detail & Related papers  (2024-07-05T17:56:38Z)
• VideoHallucer: Evaluating Intrinsic and Extrinsic Hallucinations in Large Video-Language Models [59.05674402770661]
  This work introduces VideoHallucer, the first comprehensive benchmark for hallucination detection in large video-language models (LVLMs) VideoHallucer categorizes hallucinations into two main types: intrinsic and extrinsic, offering further subcategories for detailed analysis.
  arXiv  Detail & Related papers  (2024-06-24T06:21:59Z)
• HalluciDoctor: Mitigating Hallucinatory Toxicity in Visual Instruction Data [102.56792377624927]
  hallucinations inherent in machine-generated data remain under-explored. We present a novel hallucination detection and elimination framework, HalluciDoctor, based on the cross-checking paradigm. Our method successfully mitigates 44.6% hallucinations relatively and maintains competitive performance compared to LLaVA.
  arXiv  Detail & Related papers  (2023-11-22T04:52:58Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.