無需訓練，無問題：重新思考擴散模型的無分類器指導

No Training, No Problem: Rethinking Classifier-Free Guidance for Diffusion Models

July 2, 2024

作者: Seyedmorteza Sadat, Manuel Kansy, Otmar Hilliges, Romann M. Weber

cs.AI

摘要

無分類器引導（CFG）已成為增強條件擴散模型質量的標準方法。然而，使用CFG需要訓練一個無條件模型與主要擴散模型並行，或者修改訓練程序，定期插入空條件。對於無條件模型，CFG的擴展也不明確。在本文中，我們重新審視了CFG的核心原則，並引入了一種新方法，獨立條件引導（ICG），它提供了CFG的好處，而無需任何特殊的訓練程序。我們的方法簡化了條件擴散模型的訓練過程，並且還可以應用於對任何預先訓練的條件模型進行推斷。此外，通過利用所有擴散網絡中編碼的時間步信息，我們提出了一種CFG的擴展，稱為時間步引導（TSG），可以應用於任何擴散模型，包括無條件模型。我們的引導技術易於實施，並且具有與CFG相同的採樣成本。通過廣泛的實驗，我們證明了ICG在各種條件擴散模型上與標準CFG的性能相匹配。此外，我們展示了TSG通過提高生成質量，類似於CFG的方式，而無需依賴任何條件信息。

English

Classifier-free guidance (CFG) has become the standard method for enhancing the quality of conditional diffusion models. However, employing CFG requires either training an unconditional model alongside the main diffusion model or modifying the training procedure by periodically inserting a null condition. There is also no clear extension of CFG to unconditional models. In this paper, we revisit the core principles of CFG and introduce a new method, independent condition guidance (ICG), which provides the benefits of CFG without the need for any special training procedures. Our approach streamlines the training process of conditional diffusion models and can also be applied during inference on any pre-trained conditional model. Additionally, by leveraging the time-step information encoded in all diffusion networks, we propose an extension of CFG, called time-step guidance (TSG), which can be applied to any diffusion model, including unconditional ones. Our guidance techniques are easy to implement and have the same sampling cost as CFG. Through extensive experiments, we demonstrate that ICG matches the performance of standard CFG across various conditional diffusion models. Moreover, we show that TSG improves generation quality in a manner similar to CFG, without relying on any conditional information.