<ul><li>Subject-driven generation in image generation faces challenges in data scalability and subject expansibility.</li><li>A data synthesis pipeline, utilizing in-context generation capabilities, is proposed to address these challenges.</li><li>UNO, a multi-image conditioned subject-to-image model, is introduced for controllable and consistent generation.</li><li>Experiments demonstrate the effectiveness of the proposed method in single-subject and multi-subject driven generation.</li></ul>

Less-to-More Generalization: Unlocking More Controllability by In-Context Generation

Discover more