SOTAVerified

Personalized Image Generation

Utilizes single or multiple images that contain the same subject or style, along with text prompt, to generate images that contain that subject as well as match the textual description. Includes finetuning-based methods (e.g. DreamBooth, Textual Inversion) as well as encoder-based methods (e.g. E4T, ELITE, and IP-Adapter, etc.).

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 51–58 of 58 papers

TitleStatusHype
Subject-Diffusion:Open Domain Personalized Text-to-Image Generation without Test-time Fine-tuningCode2
BLIP-Diffusion: Pre-trained Subject Representation for Controllable Text-to-Image Generation and EditingCode0
FastComposer: Tuning-Free Multi-Subject Image Generation with Localized AttentionCode2
Identity Encoder for Personalized Diffusion—0
InstantBooth: Personalized Text-to-Image Generation without Test-Time Finetuning—0
Personalized Image Generation for Color Vision Deficiency PopulationCode0
DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven GenerationCode5
An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual InversionCode5
Show:102550
← PrevPage 3 of 3Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1DreamBooth LoRA SDXL v1.0Overall (CP * PF)0.52—Unverified
2IP-Adapter ViT-G SDXL v1.0Overall (CP * PF)0.38—Unverified
3Emu2 SDXL v1.0Overall (CP * PF)0.36—Unverified
4DreamBooth SD v1.5Overall (CP * PF)0.36—Unverified
5IP-Adapter-Plus ViT-H SDXL v1.0Overall (CP * PF)0.34—Unverified
6BLIP-Diffusion SD v1.5Overall (CP * PF)0.27—Unverified
7Textual Inversion SD v1.5Overall (CP * PF)0.24—Unverified