IJCAI 2024poster0 citations
FasterVD: On Acceleration of Video Diffusion Models
Pinrui Yu, Dan Luo, Timothy Rupprecht, Lei Lu, Zhenglun Kong, Pu Zhao, Yanyu Li, Octavia Camps
Abstract
Equipped with Denoising Diffusion Probabilistic Models, video content generation has gained significant research interest recently. However, diffusion pipelines call for intensive computation and model storage, which poses challenges for their wide and efficient deployment. In this work, we address this issue by integrating LCM-LoRA to reduce the denoising steps and escalating the video generation process by frame skipping and interpolation. Our framework achieves an approximately 10× inference acceleration for high-quality realistic video generation on commonly available GPUs.
Machine Learning: ML: OptimizationMachine Learning: ML: Applications
BibTeX
@inproceedings{ijcai2024p1044,
title = {FasterVD: On Acceleration of Video Diffusion Models},
author = {Yu, Pinrui and Luo, Dan and Rupprecht, Timothy and Lu, Lei and Kong, Zhenglun and Zhao, Pu and Li, Yanyu and Camps, Octavia and Lin, Xue and Wang, Yanzhi},
booktitle = {Proceedings of the Thirty-Third International Joint Conference on
Artificial Intelligence, {IJCAI-24}},
publisher = {International Joint Conferences on Artificial Intelligence Organization},
editor = {Kate Larson},
pages = {8838--8842},
year = {2024},
month = {8},
note = {Demo Track},
doi = {10.24963/ijcai.2024/1044},
url = {https://doi.org/10.24963/ijcai.2024/1044},
}