SinFusion: Training Diffusion Models on a Single Image or Video

Yaniv Nikankin, Niv Haim, Michal Irani

Research output: Contribution to journalConference articlepeer-review


Diffusion models exhibited tremendous progress in image and video generation, exceeding GANs in quality and diversity. However, they are usually trained on very large datasets and are not naturally adapted to manipulate a given input image or video. In this paper we show how this can be resolved by training a diffusion model on a single input image or video. Our image/video-specific diffusion model (SinFusion) learns the appearance and dynamics of the single image or video, while utilizing the conditioning capabilities of diffusion models. It can solve a wide array of image/video-specific manipulation tasks. In particular, our model can learn from few frames the motion and dynamics of a single input video. It can then generate diverse new video samples of the same dynamic scene, extrapolate short videos into long ones (both forward and backward in time) and perform video upsampling. Most of these tasks are not realizable by current video-specific generation methods.

Original languageEnglish
Pages (from-to)26199-26214
Number of pages16
JournalProceedings of Machine Learning Research
StatePublished - 2023
Event40th International Conference on Machine Learning, ICML 2023 - Honolulu, United States
Duration: 23 Jul 202329 Jul 2023

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Software
  • Control and Systems Engineering
  • Statistics and Probability


Dive into the research topics of 'SinFusion: Training Diffusion Models on a Single Image or Video'. Together they form a unique fingerprint.

Cite this