Introducing Perfusion: Key-Locked Rank One Editing for Personalized Text-to-Image Models
Perfusion, a new text-to-image personalization method, has been introduced by researchers from NVIDIA and Tel Aviv University. This groundbreaking technology, which has been accepted to SIGGRAPH 2023, offers an innovative approach to personalized text-to-image models.
Perfusion excels in producing creatively personalized objects with a small additional model size of just 100KB per concept and a brief 4-minute training period. This allows significant visual alterations without losing the object’s core identity. The Key-Locking mechanism is instrumental in maintaining a consistent identity across images, while also enabling the combination of several learned concepts into one image.
In addition, Perfusion delivers flexibility at inference time, balancing visual and textual harmony with a single trained model, stretching across the entire Pareto front without extra training. This feature makes Perfusion stand out from existing models, offering a new way to portray personalized object interactions.
Perfusion has shown both qualitative and quantitative improvements over existing models, making it an ideal tool for personalizing text-to-image models. With its ease of use and impressive results, Perfusion is a must-have for anyone looking to enhance their text-to-image models.