paper-with-me

Papers

VidChain: Chain-of-Tasks with Metric-based Direct Preference Optimization for Dense Video Captioning

2025-01-12 · Ji Soo Lee, Jongha Kim, Jeehye Na, Jinyoung Park, Hyunwoo J. Kim

Despite the advancements of Video Large Language Models (VideoLLMs) in various tasks, they struggle with fine-grained temporal understanding, such as Dense Video Captioning (DVC). DVC is a complicated task of describing all events within a video while also temporally localizing them, which integrates multiple fine-grained tasks, including video segmentation, video captioning, and temporal video grounding. Previous VideoLLMs attempt to solve DVC in a single step, failing to utilize their reasoning capability. Moreover, previous training objectives for VideoLLMs do not fully reflect the evaluation metrics, therefore not providing supervision directly aligned to target tasks. To address such a problem, we propose a novel framework named VidChain comprised of Chain-of-Tasks (CoTasks) and Metric-based Direct Preference Optimization (M-DPO). CoTasks decompose a complex task into a sequence of sub-tasks, allowing VideoLLMs to leverage their reasoning capabilities more effectively. M-DPO aligns a VideoLLM with evaluation metrics, providing fine-grained supervision to each task that is well-aligned with metrics. Applied to two different VideoLLMs, VidChain consistently improves their fine-grained video understanding, thereby outperforming previous VideoLLMs on two different DVC benchmarks and also on the temporal video grounding task. Code is available at \url{https://github.com/mlvlab/VidChain}.

📄 PDF Abstract BibTeX arXiv:2501.06761

Code (1)

mlvlab/vidchain 공식 구현 pytorch

Tasks

Dense Video CaptioningVideo CaptioningVideo GroundingVideo SegmentationVideo Semantic SegmentationVideo Understanding

Similar Papers 제목 키워드 기반

RoDiF: Robust Direct Fine-Tuning of Diffusion Policies with Corrupted Human Feedback

2026-01-31 · Amitesh Vatsa, Zhixian Xie, Wanxin Jin arxiv

Diffusion policies are a powerful paradigm for robotic control, but fine-tuning them with human preferences is fundamentally challenged by the multi-step structure of the denoising process. To overcome this, we introduce…

TPO: Aligning Large Language Models with Multi-branch & Multi-step Preference Trees

2024-10-10 · Weibin Liao, Xu Chu, Yasha Wang

In the domain of complex reasoning tasks, such as mathematical reasoning, recent advancements have proposed the use of Direct Preference Optimization (DPO) to suppress output of dispreferred responses, thereby enhancing …

Mathematical Reasoning

Self-Training with Direct Preference Optimization Improves Chain-of-Thought Reasoning

2024-07-25 · Tianduo Wang, Shichen Li, Wei Lu

Effective training of language models (LMs) for mathematical reasoning tasks demands high-quality supervised fine-tuning data. Besides obtaining annotations from human experts, a common alternative is sampling from large…

Knowledge DistillationMathematical Reasoning

Reasoning Matters: Mitigate Hallucination in Multimodal Large Reasoning Models via Reasoning-Conditioned Preference Optimization

2026-05-27 · Jiawei Kong, Hao Fang, Shunxiang Liao, Jinyu Li 외 arxiv

Multimodal Large Reasoning Models introduce the reasoning paradigm, demonstrating strong capabilities on complex vision-language tasks. However, they still suffer from severe hallucinations. Existing training-based metho…

Multimodal ReasoningAnswer Generation

Uncovering the Impact of Chain-of-Thought Reasoning for Direct Preference Optimization: Lessons from Text-to-SQL

2025-02-17 · Hanbing Liu, Haoyang Li, Xiaokang Zhang, Ruotong Chen 외

Direct Preference Optimization (DPO) has proven effective in complex reasoning tasks like math word problems and code generation. However, when applied to Text-to-SQL datasets, it often fails to improve performance and c…

Code GenerationMathText to SQLText-To-SQL