Video-to-Video高清视频生成视频大模型
MS-Vid2Vid-XL旨在提升视频生成的时空连续性和分辨率，其作为Video-to-Video的第二阶段以生成720P的视频，同时还可以用于文生视频、高清视频转换等任务。其训练数据包含了精选的海量的高清视频、图像数据（最短边>=720），可以将低分辨率的视频提升到更高分辨率（1280 * 720），且其可以处理几乎任意分辨率的视频(建议16:9的宽视频)。
MS-Vid2Vid-XL aims to improve the spatiotemporal continuity and resolution of video generation. It serves as the second stage of Video-to-Video to generate 720P videos, and can also be used for various tasks such as text-to-video synthesis and high-quality video transfer. The training data includes a large collection of high-definition videos and images (with the shortest side >=720), allowing for the enhancement of low-resolution videos to higher resolutions (1280 * 720). It can handle videos of almost any resolution (preferably 16:9 aspect ratio).
Fig.1 MS-Vid2Vid-XL
模型介绍 (Introduction)
MS-Vid2Vid-XL和Video-to-Video第一阶段相同，都是基于隐空间的视频扩散模型(VLDM)，且其共享相同结构的时空UNet(ST-UNet)，其设计细节延续我们自研VideoComposer，具体可以参考其技术报告。
MS-Vid2Vid-XL and the first stage of Video-to-Video share the same underlying video latent diffusion model (VLDM). They both utilize a spatiotemporal UNet (ST-UNet) with the same structure, which is designed based on our in-house VideoComposer. For more specific details, please refer to its technical report.
依赖项（Dependency）
模型需要一下依赖才能运行
pip install modelscope
pip install xformers==0.0.21 torchsde
代码范例 (Code example)
```python
from modelscope.pipelines import pipeline
from modelscope.outputs import OutputKeys
VID_PATH: your video path
TEXT : your text description
pipe = pipeline(task="video-to-video", model='damo/Video-to-Video', modelrevision='v1.1.0', device='cuda:0')
pinput = {
'videopath': VIDPATH,
'text': TEXT
}
outputvideopath = pipe(pinput, outputvideo='./output.mp4')[OutputKeys.OUTPUT_VIDEO]
### 模型局限 (Limitation)
本**MS-Vid2Vid-XL**可能存在如下可能局限性：
- 目标距离较远时可能会存在一定的模糊，该问题可以通过输入文本来解决或缓解；
- 计算时耗大，因为需要生成720P的视频，隐空间的尺寸为(160 * 90)，单个视频计算时长>2分钟
- 目前仅支持英文，因为训练数据的原因目前仅支持英文输入
This **MS-Vid2Vid-XL** may have the following limitations:
- There may be some blurriness when the target is far away. This issue can be addressed by providing input text.
- Computation time is high due to the need to generate 720P videos. The latent space size is (160 * 90), and the computation time for a single video is more than 2 minutes.
- Currently, it only supports English. This is due to the training data, which is limited to English inputs at the moment.
## 相关论文以及引用信息 (Reference)
@article{videocomposer2023,
title={VideoComposer: Compositional Video Synthesis with Motion Controllability},
author={Wang, Xiang* and Yuan, Hangjie* and Zhang, Shiwei* and Chen, Dayou* and Wang, Jiuniu and Zhang, Yingya and Shen, Yujun and Zhao, Deli and Zhou, Jingren},
journal={arXiv preprint arXiv:2306.02018},
year={2023}
}
@inproceedings{videofusion2023,
title={VideoFusion: Decomposed Diffusion Models for High-Quality Video Generation},
author={Luo, Zhengxiong and Chen, Dayou and Zhang, Yingya and Huang, Yan and Wang, Liang and Shen, Yujun and Zhao, Deli and Zhou, Jingren and Tan, Tieniu},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
year={2023}
}
```
使用协议 (License Agreement)
我们的代码和模型权重仅可用于个人/学术研究，暂不支持商用。
Our code and model weights are only available for personal/academic research use and are currently not supported for commercial use.
评论