PulseAugur
EN
LIVE 06:58:04

Dual-End Consistency Model accelerates generative AI with improved training and sampling

By PulseAugur Editorial · [1 sources] ·

Researchers have introduced the Dual-End Consistency Model (DE-CM), a novel approach to accelerate generative models like diffusion and flow-based systems. DE-CM addresses training instability and sampling inflexibility by selecting critical sub-trajectory clusters and employing a noise-to-noisy mapping. This method achieved a state-of-the-art FID score of 1.70 in one-step generation on the ImageNet 256x256 dataset, surpassing existing one-step consistency model techniques. AI

IMPACT This model could significantly speed up generative AI tasks, making them more practical for real-world applications.

RANK_REASON The cluster contains a research paper detailing a new model and benchmark results. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Dual-End Consistency Model accelerates generative AI with improved training and sampling

COVERAGE [1]

  1. arXiv cs.CV TIER_1 English(EN) · Linwei Dong, Ruoyu Guo, Ge Bai, Zehuan Yuan, Yawei Luo, Changqing Zou ·

    Dual-End Consistency Model

    arXiv:2602.10764v3 Announce Type: replace Abstract: The slow iterative sampling nature remains a major bottleneck for the practical deployment of diffusion and flow-based generative models. While consistency models (CMs) represent a state-of-the-art distillation-based approach fo…

RELATED TOPICS