Controllable and Compositional Generation with Latent-Space Energy-Based Models

Controllable generation is one of the key requirements for successful adoption of deep generative models in real-world applications, but it still remains as a great challenge. In particular, the compositional ability to generate novel concept combinations is out of reach for most current models. In this work, we use energybased models (EBMs) to handle compositional generation over a set of attributes. To make them scalable to high-resolution image generation, we introduce an EBM in the latent space of a pre-trained generative model such as StyleGAN. We propose a novel EBM formulation representing the joint distribution of data and attributes together, and we show how sampling from it is formulated as solving an ordinary differential equation (ODE). Given a pre-trained generator, all we need for controllable generation is to train an attribute classifier. Sampling with ODEs is done efficiently in the latent space and is robust to hyperparameters. Thus, our method is simple, fast to train, and efficient to sample. Experimental results show that our method outperforms the state-of-the-art in both conditional sampling and sequential editing. In compositional generation, our method excels at zero-shot generation of unseen attribute combinations. Also, by composing energy functions with logical operators, this work is the first to achieve such compositionality in generating photo-realistic images of resolution 1024×1024. Code is available at https://github.com/NVlabs/LACE . Recent works have attempted to overcome these issues by first training an unconditional generator, and then converting it to a conditional model with a small cost [26, 20, 1, 42] . This is often achieved by discovering semantically meaningful directions in the latent space of the unconditional model. This way, one would pay the most computational cost only once for training the unconditional model. However, these approaches often struggle with compositional generation, in particular with rare 35th Conference on Neural Information Processing Systems (NeurIPS 2021).