HyperDreamBooth：用於快速個性化文本到圖像模型的HyperNetworks

摘要

個性化已成為生成式人工智慧領域中一個重要的方面，使得能夠在不同情境和風格中合成個人，同時保持對其身份的高保真度。然而，個性化過程在時間和記憶體需求方面存在固有挑戰。微調每個個性化模型需要大量 GPU 時間投資，並且以每個主題存儲一個個性化模型在存儲容量方面可能會有要求。為了克服這些挑戰，我們提出了HyperDreamBooth-一個能夠從一張人物圖像中高效生成少量個性化權重的超網絡。通過將這些權重組合到擴散模型中，再結合快速微調，HyperDreamBooth 能夠在各種情境和風格中生成一個人的臉，同時保留對多樣風格和語義修改的關鍵知識。我們的方法在大約 20 秒內實現了對臉部的個性化，比 DreamBooth 快 25 倍，比 Textual Inversion 快 125 倍，僅使用一張參考圖像，具有與 DreamBooth 相同的質量和風格多樣性。同時，我們的方法生成的模型比普通 DreamBooth 模型小 10000 倍。項目頁面：https://hyperdreambooth.github.io

English

Personalization has emerged as a prominent aspect within the field of generative AI, enabling the synthesis of individuals in diverse contexts and styles, while retaining high-fidelity to their identities. However, the process of personalization presents inherent challenges in terms of time and memory requirements. Fine-tuning each personalized model needs considerable GPU time investment, and storing a personalized model per subject can be demanding in terms of storage capacity. To overcome these challenges, we propose HyperDreamBooth-a hypernetwork capable of efficiently generating a small set of personalized weights from a single image of a person. By composing these weights into the diffusion model, coupled with fast finetuning, HyperDreamBooth can generate a person's face in various contexts and styles, with high subject details while also preserving the model's crucial knowledge of diverse styles and semantic modifications. Our method achieves personalization on faces in roughly 20 seconds, 25x faster than DreamBooth and 125x faster than Textual Inversion, using as few as one reference image, with the same quality and style diversity as DreamBooth. Also our method yields a model that is 10000x smaller than a normal DreamBooth model. Project page: https://hyperdreambooth.github.io

HyperDreamBooth：用於快速個性化文本到圖像模型的HyperNetworks

HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image Models

摘要

Support