增强定制化文本到图像生成的细节保留：一种免正则化方法
Enhancing Detail Preservation for Customized Text-to-Image Generation: A Regularization-Free Approach
AI总结：
针对定制化文本到图像生成中正则化导致细节丢失的问题，提出一种免正则化框架，通过编码器网络和新颖采样方法解决过拟合，仅需单张图像即可在半分钟内完成大规模模型定制，并保留更多细粒度细节。
摘要：
近期的文本到图像生成模型展现了生成与文本对齐且具有高保真度图像的令人印象深刻的能力。然而，生成由用户输入图像提供的新概念的图像仍然是一项具有挑战性的任务。为了解决这个问题，研究人员一直在探索各种方法来定制预训练的文本到图像生成模型。目前，大多数用于定制预训练文本到图像生成模型的现有方法都涉及使用正则化技术来防止过拟合。虽然正则化能够缓解定制的挑战，并促成在文本引导下成功创建内容，但它可能会限制模型能力，导致细节信息的丢失和性能下降。在这项工作中，我们提出了一种不使用正则化的定制化文本到图像生成新框架。具体而言，我们提出的框架由一个编码器网络和一种新颖的采样方法组成，该方法能够在不使用正则化的情况下解决过拟合问题。借助所提出的框架，我们能够在单块GPU上于半分钟内完成对大规模文本到图像生成模型的定制，且仅需用户提供一张图像。我们在实验中证明，我们提出的框架优于现有方法，并保留了更多细粒度细节。
英文摘要：
Recent text-to-image generation models have demonstrated impressive capability of generating text-aligned images with high fidelity. However, generating images of novel concept provided by the user input image is still a challenging task. To address this problem, researchers have been exploring various methods for customizing pre-trained text-to-image generation models. Currently, most existing methods for customizing pre-trained text-to-image generation models involve the use of regularization techniques to prevent over-fitting. While regularization will ease the challenge of customization and leads to successful content creation with respect to text guidance, it may restrict the model capability, resulting in the loss of detailed information and inferior performance. In this work, we propose a novel framework for customized text-to-image generation without the use of regularization. Specifically, our proposed framework consists of an encoder network and a novel sampling method which can tackle the over-fitting problem without the use of regularization. With the proposed framework, we are able to customize a large-scale text-to-image generation model within half a minute on single GPU, with only one image provided by the user. We demonstrate in experiments that our proposed framework outperforms existing methods, and preserves more fine-grained details.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
