Pith. sign in

REVIEW 6 cited by

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

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2503.19622 v1 pith:CRAHHTUF submitted 2025-03-25 cs.CV

classification cs.CV
keywords hallucinationlmmshallucinationsbenchmarkmodelmodelsreasoningvideo
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original 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.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. VADER: Adaptive Debiasing for Hallucination Mitigation in Video Large Language Models

    cs.CV 2026-08 conditional novelty 7.0 of 10

    A training-free decoding framework that adaptively reweights attention toward video tokens and erases key visual evidence per frame to suppress hallucinated predictions, achieving 72.60% accuracy on EventHallusion wit...

  2. Enhancing Video Representations with Spatiotemporal-Semantic Residual to Mitigate Hallucinations in Video Large Multimodal Models

    cs.CV 2026-01 conditional novelty 6.0 of 10

    A learned residual disruptor creates hallucination-prone negative video features, and subtracting their logits during decoding reduces hallucination in two 7B video LLMs.

  3. Position: Reasoning After Perception Means Reasoning Without Vision

    cs.CV 2025-07 conditional novelty 6.0 of 10

    Reasoning in text space cannot recover visual information that was collapsed during perception, so multimodal models need architectures that reason within the visual representation.

  4. Video-MMLU: A Massive Multi-Discipline Lecture Understanding Benchmark

    cs.CV 2025-04 conditional novelty 6.0 of 10

    Video-MMLU: a 1,065-video lecture benchmark where most AI video models score 10-50%, but text-only models answer 40% of quiz questions without video.

  5. Perception, Reason, Think, and Plan: A Survey on Large Multimodal Reasoning Models

    cs.CV 2025-05 conditional novelty 4.0 of 10

    A survey organizes multimodal reasoning research into a staged roadmap and proposes native large multimodal reasoning models that unify perception, generation, and agentic planning.

  6. FlowReasoner: Reinforcing Query-Level Meta-Agents

    cs.AI 2025-04 conditional novelty 4.0 of 10

    FlowReasoner, a reasoning meta-agent trained by R1 distillation and reinforcement learning, generates a query-specific multi-agent workflow per query and improves code benchmark accuracy over fixed workflows and direc...

Pith tools