通过迭代修正生成路径，提升图像生成模型的准确性和稳定性。
Iterative Flow Matching: Path Correction and Gradual Refinement for Enhanced Generative Modeling
- 提出迭代流匹配方法，逐步优化生成路径
- 显著减少图像生成中的幻觉现象
- 适用于各类生成模型，增强系统鲁棒性
图像生成模型已广泛应用于娱乐引导生成到逆问题求解等多种场景。然而，训练生成器是一项复杂任务，需精细调参，易产生幻觉——即生成不真实的图像。本文探讨基于流匹配的图像生成，解释并演示了流匹配为何会产生幻觉，并提出一种迭代优化流程以改进生成过程。该方法可无缝集成至几乎所有生成建模技术中，有效提升图像合成系统的性能与鲁棒性。
原文摘要 · Abstract (English)
Generative models for image generation are now commonly used for a wide variety of applications, ranging from guided image generation for entertainment to solving inverse problems. Nonetheless, training a generator is a non-trivial feat that requires fine-tuning and can lead to so-called hallucinations, that is, the generation of images that are unrealistic. In this work, we explore image generation using flow matching. We explain and demonstrate why flow matching can generate hallucinations, and propose an iterative process to improve the generation process. Our iterative process can be integrated into virtually any generative modeling technique, thereby enhancing the performance and robustness of image synthesis systems.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。