paper-with-me

홈 › Papers

DiTCtrl: Exploring Attention Control in Multi-Modal Diffusion Transformer for Tuning-Free Multi-Prompt Longer Video Generation

2024-12-24 · CVPR 2025 1 · Minghong Cai, Xiaodong Cun, Xiaoyu Li, Wenze Liu, Zhaoyang Zhang, Yong Zhang, Ying Shan, Xiangyu Yue

Sora-like video generation models have achieved remarkable progress with a Multi-Modal Diffusion Transformer MM-DiT architecture. However, the current video generation models predominantly focus on single-prompt, struggling to generate coherent scenes with multiple sequential prompts that better reflect real-world dynamic scenarios. While some pioneering works have explored multi-prompt video generation, they face significant challenges including strict training data requirements, weak prompt following, and unnatural transitions. To address these problems, we propose DiTCtrl, a training-free multi-prompt video generation method under MM-DiT architectures for the first time. Our key idea is to take the multi-prompt video generation task as temporal video editing with smooth transitions. To achieve this goal, we first analyze MM-DiT's attention mechanism, finding that the 3D full attention behaves similarly to that of the cross/self-attention blocks in the UNet-like diffusion models, enabling mask-guided precise semantic control across different prompts with attention sharing for multi-prompt video generation. Based on our careful design, the video generated by DiTCtrl achieves smooth transitions and consistent object motion given multiple sequential prompts without additional training. Besides, we also present MPVBench, a new benchmark specially designed for multi-prompt video generation to evaluate the performance of multi-prompt generation. Extensive experiments demonstrate that our method achieves state-of-the-art performance without additional training.

📄 PDF Abstract BibTeX arXiv:2412.18597

Code (1)

tencentarc/ditctrl 공식 구현 pytorch

Tasks

Video EditingVideo Generation

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Residual Connection 설명 없음
Adam 설명 없음
Diffusion Diffusion models generate samples by gradually removing noise from a signal, and their training objective can be expressed as a reweighted variational lower-bound…
Multi-Head Attention 설명 없음
Position-Wise Feed-Forward Layer 설명 없음
Label Smoothing Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…

Similar Papers 제목 키워드 기반

EditCtrl: Disentangled Local and Global Control for Real-Time Generative Video Editing

2026-02-16 · Yehonathan Litman, Shikun Liu, Dario Seyb, Nicholas Milef 외 arxiv

High-fidelity generative video editing has seen significant quality improvements by leveraging pre-trained video foundation models. However, their computational cost is a major bottleneck, as they are often designed to i…

Video Inpainting

Attention as Grounding: Exploring Textual and Cross-Modal Attention on Entities and Relations in Language-and-Vision Transformer

2022-05-01 · Findings (ACL) 2022 5 · Nikolai Ilinykh, Simon Dobnik

We explore how a multi-modal transformer trained for generation of longer image descriptions learns syntactic and semantic representations about entities and relations grounded in objects at the level of masked self-atte…

Text GenerationVisual Grounding

Attention as Grounding: Exploring Textual and Cross-Modal Attention on Entities and Relations in Language-and-Vision Transformer

2021-10-16 · ACL ARR October 2021 10 · Anonymous

We explore how a multi-modal transformer trained for generation of longer image descriptions learns syntactic and semantic representations about entities and relations grounded in objects at the level of masked self-atte…

Text GenerationVisual Grounding

MODA: MOdular Duplex Attention for Multimodal Perception, Cognition, and Emotion Understanding

2025-07-07 · Zhicheng Zhang, Wuyou Xia, Chenxi Zhao, Zhou Yan 외 arxiv

Multimodal large language models (MLLMs) recently showed strong capacity in integrating data among multiple modalities, empowered by a generalizable attention architecture. Advanced methods predominantly focus on languag…

Codewithzichao@DravidianLangTech-EACL2021: Exploring Multimodal Transformers for Meme Classification in Tamil Language

2021-04-01 · EACL (DravidianLangTech) 2021 4 · Zichao Li

This paper describes our submission to shared task on Meme Classification for Tamil Language. To address this task, we explore a multimodal transformer for meme classification in Tamil language. According to the characte…

ClassificationMeme Classification