Abstract
Diffusion-based image synthesis has attracted extensive attention recently. In particular, ControlNet that uses image-based prompts exhibits powerful capability in image tasks such as canny edge detection and generates images well aligned with these prompts. However, vanilla ControlNet generally requires extensive training of around 5000 steps to achieve a desirable control for a single task. Recent context-learning approaches have improved its adaptability, but mainly for edge-based tasks, and rely on paired examples. Thus, two important open issues are yet to be addressed to reach the full potential of ControlNet: (i) zero-shot control for certain tasks and (ii) faster adaptation for non-edge-based tasks. In this paper, we introduce a novel Meta ControlNet method, which adopts the task-agnostic meta learning technique and features a new layer freezing design. Meta ControlNet significantly reduces learning steps to attain control ability from 5000 to 1000. Further, Meta ControlNet exhibits direct zero-shot adaptability in edge-based tasks without any finetuning, and achieves control within only 100 finetuning steps in more complex non-edge tasks such as Human Pose, outperforming all existing methods. The codes is available in this https URL.
Abstract (translated)
扩散图像合成最近引起了广泛的关注。特别是,使用图像提示的控制网络在图像任务(如检测强边缘和生成与这些提示相符的图像)表现出强大的能力。然而,普通的控制网络通常需要训练约5000步才能实现对单一任务的理想控制。最近的功能扩展方法提高了其适应性,但主要针对基于边缘的任务,并且依赖于成对示例。因此,还需要解决两个重要的问题,才能发挥控制网络的全部潜力:(i)对于某些任务实现零散控制,(ii)对于非基于边缘的任务实现更快的适应。在本文中,我们引入了一种新颖的元控制网络方法,它采用任务无关的元学习技术并具有一个新的冻结层设计。元控制网络显著减少了从5000步降低到1000步的学习步骤,以实现控制能力。此外,元控制网络在边缘基于任务上直接具有零散适应能力,在更复杂的非边缘任务(如人体姿态)上,在仅需要100步微调的情况下,实现控制,超过了所有现有方法。代码可在此https:// URL中获取。
URL
https://arxiv.org/abs/2312.01255