VideoGrain: Modulating Space-Time Attention for Multi-Grained Video Editing (ICLR 2025)

Github (⭐ Star our GitHub )

If you think this dataset is helpful, please feel free to leave a star⭐️⭐️⭐️ and cite our paper:

Summary

This is the dataset proposed in our paper VideoGrain: Modulating Space-Time Attention for Multi-Grained Video Editing (ICLR 2025).

VideoGrain is a zero-shot method for class-level, instance-level, and part-level video editing.

Multi-grained Video Editing
- class-level: Editing objects within the same class (previous SOTA limited to this level)
- instance-level: Editing each individual instance to distinct object
- part-level: Adding new objects or modifying existing attributes at the part-level
Training-Free
- Does not require any training/fine-tuning
One-Prompt Multi-region Control & Deep investigations about cross/self attn
- modulating cross-attn for multi-regions control (visualizations available)
- modulating self-attn for feature decoupling (clustering are available)

Directory

data/
├── 2_cars
│   ├── 2_cars            # original videos frames
│   └── layout_masks      # layout masks subfolders (e.g., bg, left, right)
├── 2_cats
│   ├── 2_cats           
│   └── layout_masks      
├── 2_monkeys
├── badminton
├── boxer-punching
├── car
├── cat_flower
├── man_text_message
├── run_two_man
├── soap-box
├── spin-ball
├── tennis
└── wolf

Download

Automatical

Install the datasets library first, by:

pip install datasets

Then it can be downloaded automatically with

import numpy as np
from datasets import load_dataset

dataset = load_dataset("XiangpengYang/VideoGrain-dataset")

License

This dataset are licensed under the CC BY-NC 4.0 license.

Citation

@article{yang2025videograin,
  title={VideoGrain: Modulating Space-Time Attention for Multi-grained Video Editing},
  author={Yang, Xiangpeng and Zhu, Linchao and Fan, Hehe and Yang, Yi},
  journal={arXiv preprint arXiv:2502.17258},
  year={2025}
}

Contact

If you have any questions, feel free to contact Xiangpeng Yang ([email protected]).

image imagewidth (px) 512 512