This is the official repo for the following papers:

• NÜWA: Visual Synthesis Pre-training for Neural visUal World creAtion. (ECCV 2022)

• NUWA-Infinity: Autoregressive over Autoregressive Generation for Infinite Visual Synthesis. (NeurIPS 2022)

• NUWA-LIP: Language Guided Image Inpainting with Defect-free VQGAN. (CVPR 2023)

• Learning 3D Photography Videos via Self-supervised Diffusion on Single Images. (IJCAI 2023)

• NUWA-XL: Diffusion over Diffusion for eXtremely Long Video Generation. (ACL 2023)

[Project page] [Paper]

NUWA-Infinity is a generative model for infinite visual synthesis, which is defined as the task of generating arbitrarily-sized high-resolution images or long-duration videos.

NÜWA is a unified multimodal pre-trained model that can generate new or manipulate existing visual data (i.e., images and videos) for 8 visual synthesis tasks (as shown above).