Reproduce “Mise-en-Video: Fine-Grained 3D Control for Multi-Shot Video Generation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yingshuang Zou, Xiaoyu Shi, Yawen Luo, Ziang Li, Qinghe Wang, Mengfei Li, Chengfeng Zhao, Xintao Wang and 3 more
Mise-in-Video lets creators stage multi-shot videos in 3D, using oriented subject boxes and global, shot-specific, and subject-specific text. It projects layouts into the video model and routes text to relevant shots and subjects. In a layout ablation, yaw error fell from 0.4983 rad to 0.1497 rad, while pitch error rose slightly; this supports improved tested orientation control, not perfect control.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.