1 paper · 1 filter
Yuta Oshima, Daiki Miyake, Kohsei Matsutani +4
Recent text-to-image generation models have acquired the ability of multi-reference generation and editing; that is, to inherit the appearance of subjects from multiple reference i…