paper-with-me

Papers

Seq2Time: Sequential Knowledge Transfer for Video LLM Temporal Grounding

2024-11-25 · CVPR 2025 1 · Andong Deng, Zhongpai Gao, Anwesa Choudhuri, Benjamin Planche, Meng Zheng, Bin Wang, Terrence Chen, Chen Chen, Ziyan Wu

Temporal awareness is essential for video large language models (LLMs) to understand and reason about events within long videos, enabling applications like dense video captioning and temporal video grounding in a unified system. However, the scarcity of long videos with detailed captions and precise temporal annotations limits their temporal awareness. In this paper, we propose Seq2Time, a data-oriented training paradigm that leverages sequences of images and short video clips to enhance temporal awareness in long videos. By converting sequence positions into temporal annotations, we transform large-scale image and clip captioning datasets into sequences that mimic the temporal structure of long videos, enabling self-supervised training with abundant time-sensitive data. To enable sequence-to-time knowledge transfer, we introduce a novel time representation that unifies positional information across image sequences, clip sequences, and long videos. Experiments demonstrate the effectiveness of our method, achieving a 27.6% improvement in F1 score and 44.8% in CIDEr on the YouCook2 benchmark and a 14.7% increase in recall on the Charades-STA benchmark compared to the baseline.

📄 PDF Abstract BibTeX arXiv:2411.16932

Code (0)

등록된 구현이 없습니다.

Tasks

Dense Video CaptioningTransfer LearningVideo CaptioningVideo Grounding

Methods 이 논문이 사용한 방법론

CLIP Contrastive Language-Image Pre-training (CLIP), consisting of a simplified version of ConVIRT trained from scratch, is an efficient method of image representation learning…

Similar Papers 제목 키워드 기반

AdapNet: Adaptability Decomposing Encoder-Decoder Network for Weakly Supervised Action Recognition and Localization

2019-11-27 · Xiao-Yu Zhang, Changsheng Li, Haichao Shi, Xiaobin Zhu 외

The point process is a solid framework to model sequential data, such as videos, by exploring the underlying relevance. As a challenging problem for high-level video understanding, weakly supervised action recognition an…

Action ClassificationAction RecognitionDecoderTemporal Localization+3

Transfer of Temporal Logic Formulas in Reinforcement Learning

2019-09-10 · Zhe Xu, Ufuk Topcu

Transferring high-level knowledge from a source task to a target task is an effective way to expedite reinforcement learning (RL). For example, propositional logic and first-order logic have been used as representations …

reinforcement-learningReinforcement LearningReinforcement Learning (RL)Transfer Learning

Learning Transferable Spatiotemporal Representations from Natural Script Knowledge

2022-09-30 · CVPR 2023 1 · Ziyun Zeng, Yuying Ge, Xihui Liu, Bin Chen 외

Pre-training on large-scale video data has become a common recipe for learning transferable spatiotemporal representations in recent years. Despite some progress, existing methods are mostly limited to highly curated dat…

DescriptiveRepresentation LearningVideo Understanding

Lifelong Reinforcement Learning with Temporal Logic Formulas and Reward Machines

2021-11-18 · Xuejing Zheng, Chao Yu, Chen Chen, Jianye Hao 외

Continuously learning new tasks using high-level ideas or knowledge is a key capability of humans. In this paper, we propose Lifelong reinforcement learning with Sequential linear temporal logic formulas and Reward Machi…

Lifelong learningreinforcement-learningReinforcement LearningReinforcement Learning (RL)+1

From Pretrain to Pain: Adversarial Vulnerability of Video Foundation Models Without Task Knowledge

2025-11-10 · Hui Lu, Yi Yu, Song Xia, Yiming Yang 외 arxiv

Large-scale Video Foundation Models (VFMs) has significantly advanced various video-related tasks, either through task-specific models or Multi-modal Large Language Models (MLLMs). However, the open accessibility of VFMs…

Contrastive LearningAdversarial Attack