Spotlight Attention: Robust Object-Centric Learning With a Spatial Locality Prior
This work addresses the challenge of robust unsupervised object-centric learning for computer vision, representing an incremental improvement over existing methods.
The paper tackled the problem of weak spatial continuity in object-centric vision models by incorporating a spatial-locality prior, resulting in significant improvements in object segmentation on synthetic and real-world datasets, with enhanced robustness to hyperparameters.
The aim of object-centric vision is to construct an explicit representation of the objects in a scene. This representation is obtained via a set of interchangeable modules called \emph{slots} or \emph{object files} that compete for local patches of an image. The competition has a weak inductive bias to preserve spatial continuity; consequently, one slot may claim patches scattered diffusely throughout the image. In contrast, the inductive bias of human vision is strong, to the degree that attention has classically been described with a spotlight metaphor. We incorporate a spatial-locality prior into state-of-the-art object-centric vision models and obtain significant improvements in segmenting objects in both synthetic and real-world datasets. Similar to human visual attention, the combination of image content and spatial constraints yield robust unsupervised object-centric learning, including less sensitivity to model hyperparameters.