DisCont: Self-Supervised Visual Attribute Disentanglement Using Context Vectors

Sarthak Bhagat, Vishaal Udandarao, Shagun Uppal, Saket Anand

Abstract: Disentangling underlying feature attributes within an image with no prior supervision is a challenging task. Models that can disentangle attributes well, provide greater interpretability and control. In this paper, we propose a self-supervised framework: DisCont to disentangle multiple attributes by exploiting structural inductive biases within images. Motivated by a recent surge in contrastive learning frameworks, our model bridges the gap between self-supervised contrastive learning algorithms and unsupervised disentanglement. We evaluate the efficacy of our approach, qualitatively and quantitatively, on four benchmark datasets. The code is available at https://github.com/ sarthak268/DisCont .