On Inference Stability for Diffusion Models
DOI:
https://doi.org/10.1609/aaai.v38i13.29359Keywords:
ML: Deep Generative Models & AutoencodersAbstract
Denoising Probabilistic Models (DPMs) represent an emerging domain of generative models that excel in generating diverse and high-quality images. However, most current training methods for DPMs often neglect the correlation between timesteps, limiting the model's performance in generating images effectively. Notably, we theoretically point out that this issue can be caused by the cumulative estimation gap between the predicted and the actual trajectory. To minimize that gap, we propose a novel sequence-aware loss that aims to reduce the estimation gap to enhance the sampling quality. Furthermore, we theoretically show that our proposed loss function is a tighter upper bound of the estimation loss in comparison with the conventional loss in DPMs. Experimental results on several benchmark datasets including CIFAR10, CelebA, and CelebA-HQ consistently show a remarkable improvement of our proposed method regarding the image generalization quality measured by FID and Inception Score compared to several DPM baselines. Our code and pre-trained checkpoints are available at https://github.com/VinAIResearch/SA-DPM.Downloads
Published
2024-03-24
How to Cite
Nguyen, V., Vu, G., Nguyen Thanh, T., Than, K., & Tran, T. (2024). On Inference Stability for Diffusion Models. Proceedings of the AAAI Conference on Artificial Intelligence, 38(13), 14449-14456. https://doi.org/10.1609/aaai.v38i13.29359
Issue
Section
AAAI Technical Track on Machine Learning IV