I2VGen-XL: High-Quality Image-to-Video Synthesis
via Cascaded Diffusion Models
Video synthesis has recently made remarkable strides benefiting from the rapid development of diffusion models. However, it still encounters challenges in terms of semantic accuracy, clarity and spatio-temporal continuity. They primarily arise from the scarcity of well-aligned text-video data and the complex inherent structure of videos, making it difficult for the model to simultaneously ensure semantic and qualitative excellence. In this report, we propose a cascaded I2VGen-XL approach that enhances model performance by decoupling these two factors and ensures the alignment of the input data by utilizing static images as a form of crucial guidance. I2VGen-XL consists of two stages: i) the base stage guarantees coherent semantics and preserves content from input images by using two hierarchical encoders, and ii) the refinement stage enhances the video's details by incorporating an additional brief text and improves the resolution to 1280x720. To improve the diversity, we collect around 35 million single-shot text-video pairs and 6 billion text-image pairs to optimize the model. By this means, I2VGen-XL can simultaneously enhance the semantic accuracy, continuity of details and clarity of generated videos. Through extensive experiments, we have investigated the underlying principles of I2VGen-XL and compared it with current top methods, which can demonstrate its effectiveness on diverse data. The source code and models will be publicly available here.
A dog in a suit and tie faces the camera.
Papers were floating in the air on a table in the library.
In a rice field , a girl walks toward the eye of the storm with her back to the camera.
A monster was sitting on a house on the side of the street , with its head pointed at the people in the street.
A close-up of a parrot.
A small boat floating on the calm water, 3D cartoon.
Waterfall in the forest.
Deer in the sunset.
Deep-sea shark.
A man's face has a flashing circuit light path.
A dreamy garden with a city in the distance.
A zombie old man with white hair stood on the podium with his hands raised.
Goldfish in Chinese ink painting style for a few days.
A song in space in space.
A cute little girl smiling at the camera, 2D culture.
We have the opportunity waiting for you.
If you are seeking an exhilarating challenge and the chance to collaborate with AIGC and large-scale pretraining, then you have come to the right place. We are searching for talented, motivated, and imaginative researchers to join our team. If you are interested, please don't hesitate to send us your resume via email zhangjin.zsw@alibaba-inc.com
References
@article{2023i2vgenxl,
title={I2VGen-XL: High-Quality Image-to-Video Synthesis via Cascaded Diffusion Models},
author={Zhang, Shiwei* and Wang, Jiayu* and Zhang, Yingya* and Zhao, Kang and Yuan, Hangjie and Qing, Zhiwu and Wang, Xiang and Zhao, Deli and Zhou, Jingren},
booktitle={arXiv preprint arXiv:2311.04145},
year={2023}
}
@article{2023videocomposer,
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},
booktitle={arXiv preprint arXiv:2306.02018},
year={2023}
}