BM-6M

[](https://arxiv.org/abs/2506.03107)
[](https://boese0601.github.io/bytemorph/)
[](https://huggingface.co/datasets/ByteDance-Seed/BM-Bench)
[](https://huggingface.co/datasets/ByteDance-Seed/BM-6M-Demo)
[](https://huggingface.co/datasets/ByteDance-Seed/BM-6M)
[](https://huggingface.co/spaces/Boese0601/ByteMorpher-Demo)
[](https://huggingface.co/ByteDance-Seed/BM-Model)
[](https://github.com/ByteDance-Seed/BM-code)

# Dataset Card for ByteMorph-6M
The task of editing images to reflect non-rigid motions, such as changes in camera viewpoint, object deformation, human articulation, or complex interactions, represents a significant yet underexplored frontier in computer vision. Current methodologies and datasets often concentrate on static imagery or rigid transformations, thus limiting their applicability to expressive edits involving dynamic movement. To bridge this gap, we present ByteMorph, a substantial benchmark specifically created for instruction-based image editing focused on non-rigid motions. This dataset card contains the example training data subset and instructions for ByteMorph-6M.

## Dataset Details

Original videos are generated by [Seaweed](https://seaweed.video/) and sampled into frames as source-target image editing pairs. These frames are further filtered and captioned by VLM. For visualization of a subset of the whole dataset, please visit [this repo](https://huggingface.co/datasets/ByteDance-Seed/BM-6M-Demo).

## Intended use
Primary intended uses: The primary use of ByteMorph is research on text-to-image and instruction-based image editing.

Primary intended users: The model's primary intended users are researchers and hobbyists in computer vision, image generation, image processing, and AIGC.

## Dataset Structure

```bash
BM-6M
|----subset-1
|----sample_frames # extracted first and last frames from the video
|----batch_0.tar
|----batch_1.tar
|----...
|----sample_multi_frames # extracted multi frames from the video
|----batch_0.tar
|----batch_1.tar
|----...
|----subset-2
|----subset-3
|----...
|----subset-9
```

### How to use ByteMorph-6M
Simply download this dataset with [git-lfs](https://github.com/git-lfs/git-lfs/blob/main/INSTALLING.md). You can also download the subset of the whole dataset.
```bash
git lfs clone https://huggingface.co/datasets/ByteDance-Seed/BM-6M
```

## Bibtex citation

```bibtex
@article{chang2025bytemorph,
title={ByteMorph: Benchmarking Instruction-Guided Image Editing with Non-Rigid Motions},
author={Chang, Di and Cao, Mingdeng and Shi, Yichun and Liu, Bo and Cai, Shengqu and Zhou, Shijie and Huang, Weilin and Wetzstein, Gordon and Soleymani, Mohammad and Wang, Peng},
journal={arXiv preprint arXiv:2506.03107},
year={2025}
}
```

[](https://arxiv.org/abs/2506.03107)
[](https://boese0601.github.io/bytemorph/)
[](https://huggingface.co/datasets/ByteDance-Seed/BM-Bench)
[](https://huggingface.co/datasets/ByteDance-Seed/BM-6M-Demo)
[](https://huggingface.co/datasets/ByteDance-Seed/BM-6M)
[](https://huggingface.co/spaces/Boese0601/ByteMorpher-Demo)
[](https://huggingface.co/ByteDance-Seed/BM-Model)
[](https://github.com/ByteDance-Seed/BM-code)

# ByteMorph-6M 数据集卡片
旨在编辑图像以呈现非刚性运动（non-rigid motions，如相机视角变化、物体形变、人体关节运动或复杂交互）的任务，是计算机视觉领域中一项极具价值却尚未得到充分探索的前沿方向。当前的研究方法与数据集往往聚焦于静态图像或刚性变换（rigid transformations），这限制了其在涉及动态运动的富有表现力的图像编辑任务中的应用。为填补这一空白，我们推出了ByteMorph：一款专为聚焦非刚性运动（non-rigid motions）的指令式图像编辑（instruction-based image editing）任务打造的大规模基准测试集。本数据集卡片包含ByteMorph-6M的示例训练数据子集与使用说明。

## 数据集详情
原始视频由[Seaweed](https://seaweed.video/)生成，并采样为帧对作为源-目标图像编辑样本。随后通过视觉语言模型（Visual Language Model，简称VLM）对这些帧进行进一步筛选与字幕标注。如需查看全量数据集的子集可视化效果，请访问[此仓库](https://huggingface.co/datasets/ByteDance-Seed/BM-6M-Demo)。

## 预期用途与目标用户
### 核心预期用途
ByteMorph的核心用途为面向文本到图像生成（text-to-image）与指令式图像编辑的研究。
### 目标用户群体
该基准集的目标用户为计算机视觉、图像生成、图像处理以及生成式人工智能内容（Artificial Intelligence Generated Content，简称AIGC）领域的研究人员与爱好者。

## 数据集结构
bash
BM-6M
|----subset-1
|----sample_frames # 从视频中提取的首帧与末帧
|----batch_0.tar
|----batch_1.tar
|----...
|----sample_multi_frames # 从视频中提取的多帧
|----batch_0.tar
|----batch_1.tar
|----...
|----subset-2
|----subset-3
|----...
|----subset-9

## ByteMorph-6M 使用方法
### 数据集下载方式
你可以通过[git-lfs](https://github.com/git-lfs/git-lfs/blob/main/INSTALLING.md)直接下载该数据集，也可仅下载全量数据集的子集。
bash
git lfs clone https://huggingface.co/datasets/ByteDance-Seed/BM-6M

## 参考文献引用
bibtex
@article{chang2025bytemorph,
title={ByteMorph: Benchmarking Instruction-Guided Image Editing with Non-Rigid Motions},
author={Chang, Di and Cao, Mingdeng and Shi, Yichun and Liu, Bo and Cai, Shengqu and Zhou, Shijie and Huang, Weilin and Wetzstein, Gordon and Soleymani, Mohammad and Wang, Peng},
journal={arXiv preprint arXiv:2506.03107},
year={2025}
}