Earlier Research: Efficient Generative Modeling

One-step diffusion generation through score and distribution distillation

My earlier research focused on efficient generative modeling, developing score-based distillation methods for training one-step diffusion generators, including SiD and SiDA.

Publications

Adversarial Score identity Distillation: Rapidly Surpassing the Teacher in One Step

2025-02-15 Mingyuan Zhou, Huangjie Zheng, Yi Gu, Zhendong Wang, Hai Huang ICLR 2025

A one-step adversarial distillation method for diffusion models that improves both sample quality and distillation efficiency.

Paper

Score identity Distillation: Exponentially Fast Distillation of Pretrained Diffusion Models for One-Step Generation

2024-04-15 Mingyuan Zhou, Huangjie Zheng, Zhendong Wang, Mingzhang Yin, Hai Huang ICML 2024

A data-free method that distills the generative capabilities of pretrained diffusion models into a single-step generator.

Paper