Lune

CVPR2025顶会

LoRACLR: Contrastive Adaptation for Customization of Diffusion Models

Enis Simsar, Thomas Hofmann, Federico Tombari, Pinar Yanardag

2025年份
7顶会引用

摘要

<PITT> & <MARGOT>, inside a futuristic spaceship, sci-fi realism… <MESSI> & <TAYLOR> & <LEBRON> , on a seashore, wearing denim jackets, in <COMIC> style… CONCEPTS Figure 1. High-Fidelity Multi-Concept Image Generation. Examples illustrating LoRACLR's ability to generate unified scenes with multiple distinct characters and styles across diverse settings. Each scene demonstrates LoRACLR's capability to combine varied concepts seamlessly, preserving the original identities of each character, as seen in concepts.

问问这篇 Paper

智能体会读完全文。

Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。

可以从这些问题问起

智能体调用

Luneget_paper_fulltext

在 Lune 里问

免费开始,无需绑卡

引用它的顶会 Paper7

问问它们各自怎么用它

它引用的顶会 Paper19

相关 Paper

黄昏的海面,两侧是细线勾勒的悬崖