Multi-Concept Customization of Text-to-Image Diffusion

1 CMU    2Tsinghua University    3Adobe Research

Code Paper Project Gallery Slides CustomConcept101

Please refer to our code for details regarding dataset download and evaluation. Sample images of each concept in the Dataset are shown below.

Barn


Lighthouse


Castle


Canal Scene


Garden


Waterfall


Sculpture


Sculpture


Houseplant


Houseplant


Houseplant


Vase


Wooden-pot


Lamp


Cup


Helmet


Sunglasses


Sunglasses


Backpack


Glasses


Jacket


Jacket


Purse


Purse


Purse


Purse


Earring


Ring


Shoes


Shoes


Key-chain


Bear


Cup


Cup


Corkscrew


Headphone


Headphone


Book


book


Bottle


Cow Plushie


Bunny


Panda Plushie


Lobster Plushie


Penguin Plushie


Dice Plushie


Plushie


Unicorn Plushie


Tortoise Plushie


Happy Sad Plushie


Sad Plushie


Plushie


Teddybear


Music Instrument


Music Instrument


Music Instrument


Sofa


Sofa


Chair


Chair


Chair


Bike


Car


Car


Car


Car


Car


Car


Car


Car


Car


Car


Car


Motorbike


Tank


Dog


Dog


Dog


Dog


Cat


Cat


Cat


Cat


Cat


Cat


Kids' Table-Chair


Pokemon Toy


Toy Gnome


Toy Unicorn


Dish


Dish


Person


Person


Flower


Flower


Guitar Amplifier


Tuner


Action Figure


Action Figure


Action Figure



Acknowledgements

We are grateful to Sheng-Yu Wang, Songwei Ge, Daohan Lu, Ruihan Gao, Roni Shechtman, Avani Sethi, Yijia Wang, and Shagun Uppal for helping with the dataset collection.