PaMi-VDPO: Mitigating Video Hallucinations by Prompt-Aware Multi-Instance Video Preference Learning
Direct Preference Optimization (DPO) helps reduce hallucinations in Video Multimodal Large Language Models (VLLMs), but its reliance on offline preference data limits adaptability and fails to capture true video-response misalignment. We propose Video Direct Preference Optimization (VDPO), an online preference learning framework that eliminates the need for preference annotation by leveraging video augmentations to generate rejected samples while keeping responses fixed. However, selecting effective augmentations is non-trivial, as some clips may be semantically identical to the original under specific prompts, leading to false rejections and disrupting alignment. To address this, we introduce Prompt-aware Multi-instance Learning VDPO (PaMi-VDPO), which selects augmentations based on prompt context. Instead of a single rejection, we construct a candidate set of augmented clips and apply a close-to-far selection strategy, initially ensuring all clips are semantically relevant while then prioritizing the most prompt-aware distinct clip. This allows the model to better capture meaningful visual differences, mitigating hallucinations, while avoiding false rejections, and improving alignment. PaMi-VDPOseamlessly integrates into existing VLLMs without additional parameters, GPT-4/human supervision. With only 10k SFT data, it improves the base model by 5.3% on VideoHallucer, surpassing GPT-4o, while maintaining stable performance on general video benchmarks.
Code (0)
등록된 구현이 없습니다.
Methods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Vision-Driven Prompt Optimization for Large Language Models in Multimodal Generative Tasks
Vision generation remains a challenging frontier in artificial intelligence, requiring seamless integration of visual understanding and generative capabilities. In this paper, we propose a novel framework, Vision-Driven …
Image GenerationMitigating Hallucination in VideoLLMs via Temporal-Aware Activation Engineering
Multimodal large language models (MLLMs) have achieved remarkable progress in video understanding.However, hallucination, where the model generates plausible yet incorrect outputs, persists as a significant and under-add…
HallucinationVIPAMIN: Visual Prompt Initialization via Embedding Selection and Subspace Expansion
In the era of large-scale foundation models, fully fine-tuning pretrained networks for each downstream task is often prohibitively resource-intensive. Prompt tuning offers a lightweight alternative by introducing tunable…
Visual Prompt TuningCPR: Mitigating Large Language Model Hallucinations with Curative Prompt Refinement
Recent advancements in large language models (LLMs) highlight their fluency in generating responses to diverse prompts. However, these models sometimes generate plausible yet incorrect ``hallucinated" facts, undermining …
Listening to Patients: A Framework of Detecting and Mitigating Patient Misreport for Medical Dialogue Generation
Medical Dialogue Systems aim to provide automated healthcare support through patient-agent conversations. Previous efforts typically regard patients as ideal users -- one who accurately and consistently reports their hea…
Dialogue GenerationHallucinationMisconceptionsResponse Generation