Logo image
Back
ACT-Diffusion: Efficient Adversarial Consistency Training for One-Step Diffusion Models
Conference proceeding   Open access

ACT-Diffusion: Efficient Adversarial Consistency Training for One-Step Diffusion Models

Fei Kong, Jinhao Duan, Lichao Sun, Hao Cheng, Renjing Xu, Hengtao Shen, Xiaofeng Zhu, Xiaoshuang Shi and Kaidi Xu
2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp 8890-8899
16 Jun 2024
DOI: https://doi.org/10.1109/CVPR52733.2024.00849
url
http://arxiv.org/abs/2311.14097View
SubmittedarXiv.org - Non-exclusive license to distribute,  Open

Abstract

Adaptation models Costs Diffusion models Image synthesis Training Upper bound Noise Reduction
Though diffusion models excel in image generation, their step-by-step denoising leads to slow generation speeds. Consistency training addresses this issue with single-step sampling but often produces lower-quality generations and requires high training costs. In this paper, we show that optimizing consistency training loss minimizes the Wasserstein distance between target and generated distributions. As timestep increases, the upper bound accumulates previous consistency training losses. Therefore, larger batch sizes are needed to reduce both current and accumulated losses. We propose Adversarial Consistency Training (ACT), which directly minimizes the Jensen-Shannon (JS) divergence between distributions at each timestep using a discriminator. Theoretically, ACT enhances generation quality, and convergence. By incorporating a discriminator into the consistency training framework, our method achieves improved FID scores on CIFAR10 and ImageNet 64×64 and LSUN Cat 256 ×256 datasets, retains zero-shot image inpainting capabilities, and uses less than 1/6 of the original batch size and fewer than 1/2 of the model parameters and training steps compared to the baseline method, this leads to a substantial reduction in resource consumption. Our code is available: https://github.com/kong13661/ACT

Metrics

14 Record Views
2 citations in Scopus

Details

Logo image