ReDirector: Creating Any-Length Video Retakes with Rotary Camera Encoding
Byeongjun Park, Byung-Hoon Kim, Hyungjin Chung, Jong Chul
摘要
We present ReDirector, a novel camera-controlled video retake generation method for dynamically captured variable-length videos. In particular, we rectify a common misuse of RoPE in previous works by aligning the spatiotemporal positions of the input video and the target retake. Moreover, we introduce Rotary Camera Encoding (RoCE), a camera-conditioned RoPE phase shift that captures and integrates multi-view relationships within and across the input and target videos. By integrating camera conditions into RoPE, our method generalizes to out-of-distribution camera trajectories and video lengths, yielding improved dynamic object localization and static background preservation. Extensive experiments further demonstrate significant improvements in camera controllability, geometric consistency, and video quality across various trajectories and lengths.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
它引用的顶会 Paper46
- BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and GenerationJunnan Li, Dongxu Li, Caiming Xiong, Steven C. H. HoiICML 2022 · 被引用 6,549 次
- Scalable Diffusion Models with TransformersWilliam Peebles, Saining XieICCV 2023 · 被引用 5,568 次
- Scaling Rectified Flow Transformers for High-Resolution Image SynthesisPatrick Esser, Sumith Kulal, Andreas Blattmann, Rahim Entezari 等ICML 2024 · 被引用 3,620 次
- Dynamic View Synthesis from Dynamic Monocular VideoChen Gao, Ayush Saraf, Johannes Kopf, Jia-Bin HuangICCV 2021 · 被引用 522 次
- CAT3D: Create Anything in 3D with Multi-View Diffusion ModelsRuiqi Gao, Aleksander Holynski, Philipp Henzler, Arthur Brussee 等NeurIPS 2024 · 被引用 490 次
相关 Paper
- ReRoPE: Repurposing RoPE for Relative Camera ControlChunyang Li, Yuanbo Yang, Jiahao Shao, Hongyu Zhou 等SIGGRAPH 2026 · 被引用 2 次
- MultiShotMaster: A Controllable Multi-Shot Video Generation FrameworkQinghe Wang, Xiaoyu Shi, Baolu Li, Weikang Bian 等CVPR 2026 · 被引用 33 次
- VRoPE: Rotary Position Embedding for Video Large Language ModelsZikang Liu, Longteng Guo, Yepeng Tang, Tongtian Yue 等EMNLP 2025 · 被引用 1 次
- VideoRoPE: What Makes for Good Video Rotary Position Embedding?Xilin Wei, Xiaoran Liu, Yuhang Zang, Xiaoyi Dong 等ICML 2025
- A Circular Argument: Does RoPE need to be Equivariant for Vision?Chase van de Geijn, Timo Lüddecke, Polina Turishcheva, Alexander S. EckerNeurIPS 2025 · 被引用 6 次
