LKGD

Codes for Latent Knowledge-Guided Video Diffusion for Scientific Phenomena Generation from a Single Initial Frame

cd svd-train-main/train_models
bash train_scripts/train_svd_lora.sh

Project Page

https://qinglongcao.xyz/LKGD-Diffusion.github.io/

Datasets and Data Preparation

The Utilized Datasets (Fluid & Typhoon)

Google Driver Datasets

BibTeX

@inproceedings{cao2026latent,
  title={Latent knowledge-guided video diffusion for scientific phenomena generation from a single initial frame},
  author={Cao, Qinglong and Li, Xirui and Wang, Ding and Ma, Chao and Chen, Yuntian and Yang, Xiaokang},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={40},
  number={4},
  pages={2625--2633},
  year={2026}
}