Inductive Biases for Object-Centric Representations in the Presence of Complex Textures

Samuele Papa, Ole Winther, Andrea Dittadi

Keywords: deep learning, representation learning, structured representations, object-centric learning, complex textures

TL;DR: When objects have complex textures, object-centric methods that use a single module to reconstruct both the shape and visual appearance of each object tend to more easily learn more useful representations and achieve better object separation.

Abstract: Understanding which inductive biases could be helpful for the unsupervised learning of object-centric representations of natural scenes is challenging. In this paper, we systematically investigate the performance of two models on datasets where neural style transfer was used to obtain objects with complex textures while still retaining ground-truth annotations. We find that by using a single module to reconstruct both the shape and visual appearance of each object, the model learns more useful representations and achieves better object separation. In addition, we observe that adjusting the latent space size is insufficient to improve segmentation performance. Finally, the downstream usefulness of the representations is significantly more strongly correlated with segmentation quality than with reconstruction accuracy.