Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

DomainGallery: Few-shot Domain-driven Image Generation by Attribute-centric Finetuning

About

The recent progress in text-to-image models pretrained on large-scale datasets has enabled us to generate various images as long as we provide a text prompt describing what we want. Nevertheless, the availability of these models is still limited when we expect to generate images that fall into a specific domain either hard to describe or just unseen to the models. In this work, we propose DomainGallery, a few-shot domain-driven image generation method which aims at finetuning pretrained Stable Diffusion on few-shot target datasets in an attribute-centric manner. Specifically, DomainGallery features prior attribute erasure, attribute disentanglement, regularization and enhancement. These techniques are tailored to few-shot domain-driven generation in order to solve key issues that previous works have failed to settle. Extensive experiments are given to validate the superior performance of DomainGallery on a variety of domain-driven generation scenarios. Codes are available at https://github.com/Ldhlwh/DomainGallery.

Yuxuan Duan, Yan Hong, Bo Zhang, Jun Lan, Huijia Zhu, Weiqiang Wang, Jianfu Zhang, Li Niu, Liqing Zhang• 2024

Related benchmarks

TaskDatasetResultRank
Few-shot Image GenerationSunglasses 10-shot
FID43.1
36
Few-shot Image GenerationBabies 10-shot
FID58.86
35
Few-shot Image GenerationAFHQ-Dog 10-shot
FID123.5
34
Few-shot Image GenerationAFHQ-Wild 10-shot
FID65.32
34
Few-shot Image GenerationMetFaces 10-shot
FID60.38
34
Few-shot Image GenerationAFHQ-Cat 10-shot
FID77.15
34
Few-shot Image GenerationSketches 10-shot
FID44.86
18
Intra-category image generationCUFS sketches
FID44.86
4
Intra-category image generationVan Gogh houses
KID32.2
4
Intra-category image generationFFHQ sunglasses
FID43.1
4
Showing 10 of 12 rows

Other info

Code

Follow for update