Learning to Manipulate Individual Objects in an Image

We describe a method to train a generative model with latent factors that are (approximately) independent and localized. This means that perturbing the latent variables affects only local regions of the synthesized image, corresponding to objects. Unlike other unsupervised generative models, ours enables object-centric manipulation, without requiring object-level annotations, or any form of annotation for that matter. The key to our method is the combination of spatial disentanglement, enforced by a Contextual Information Separation loss, and perceptual cycleconsistency, enforced by a loss that penalizes changes in the image partition in response to perturbations of the latent factors. We test our method's ability to allow independent control of spatial and semantic factors of variability on existing datasets, and also introduce two new ones which highlight the limitations of current methods. 1 * Equal contribution. † Work is done during the author's visit at UCLA.