Semantica: An Adaptable Image-Conditioned Diffusion Model
By: Manoj Kumar, Neil Houlsby, Emiel Hoogeboom Posted: May 24, 2024
arXiv:2405.14857v1 Announce Type: cross
Abstract: We investigate the task of adapting image generative models to different datasets without finetuneing. To this end, we introduce Semantica, an image-conditioned diffusion model capable of generating images based on the semantics of a conditioning image. Semantica is trained exclusively on web-scale image pairs, that is it receives a random image from a webpage as conditional input and models another random image from the same webpage. Our experiments highlight the expressivity of pretrained image encoders and necessity of semantic-based data filtering in achieving high-quality image generation. Once trained, it can adaptively generate new images from a dataset by simply using images from that dataset as input. We study the transfer properties of Semantica on ImageNet, LSUN Churches, LSUN Bedroom and SUN397.
Provided by: