For example, hats and sunglasses are placed on incorrect objects. We introduce a novel method T oMeto address these challenges.

While supervised methods are effective, they require extensive labeled data for training. Recently, self-supervised foundation models have been introduced to these tasks showing comparable results to supervised methods.

We start from an intuition that many complex visual data, e.g., videos and 3D scenes, are represented

In this work, we present MuDI, a novel framework that enables multi-subject personalization by effectively decoupling identities from multiple subjects.

Despite their demonstrated utility for NLP, multimodal counterfactual examples have been relatively unexplored due to the difficulty of creating paired image-text data with minimal counterfactual changes. To address this challenge, we introduce a scalable framework for automatic generation of counterfactual examples using text-to-image diffusion models.

We then propose to employ layout, a very abstract and easy-to-obtain prior, as the spatial guidance for subject arrangement.