Self-Corrected Flow Distillation for Consistent One-Step and Few-Step Text-to-Image Generation
This addresses the computational inefficiency of flow-matching models for researchers and practitioners in generative AI, though it appears incremental as it builds on existing frameworks.
The paper tackles the problem of reducing sampling steps in flow-matching generative models for text-to-image generation, achieving consistent quality in both few-step and one-step sampling with superior quantitative and qualitative results on CelebA-HQ and COCO benchmarks.
Flow matching has emerged as a promising framework for training generative models, demonstrating impressive empirical performance while offering relative ease of training compared to diffusion-based models. However, this method still requires numerous function evaluations in the sampling process. To address these limitations, we introduce a self-corrected flow distillation method that effectively integrates consistency models and adversarial training within the flow-matching framework. This work is a pioneer in achieving consistent generation quality in both few-step and one-step sampling. Our extensive experiments validate the effectiveness of our method, yielding superior results both quantitatively and qualitatively on CelebA-HQ and zero-shot benchmarks on the COCO dataset. Our implementation is released at https://github.com/VinAIResearch/SCFlow