Abstract Algorithms based on deep network models are being used for many pattern recognition and decision-making tasks in robotics and AI. Training these models requires a large labeled dataset and considerable computational resources, which are not readily available in many domains. Also, it is difficult to explore the internal representations and reasoning mechanisms of these models. As a step towards addressing the underlying knowledge representation, reasoning, and learning challenges, the architecture described in this paper draws inspiration from research in cognitive systems. As a motivating example, we consider an assistive robot trying to reduce clutter in any given scene by reasoning about the occlusion of objects and stability of object configurations in an image of the scene. In this context, our architecture incrementally learns and revises a grounding of the spatial relations between objects and uses this grounding to extract spatial information from input images. Non-monotonic logical reasoning with this information and incomplete commonsense domain knowledge is used to make decisions about stability and occlusion. For images that cannot be processed by such reasoning, regions relevant to the tasks at hand are automatically identified and used to train deep network models to make the desired decisions. Image regions used to train the deep networks are also used to incrementally acquire previously unknown state constraints that are merged with the existing knowledge for subsequent reasoning. Experimental evaluation performed using simulated and real-world images indicates that in comparison with baselines based just on deep networks, our architecture improves reliability of decision making and reduces the effort involved in training data-driven deep network models.

Our architecture uses non-monotonic logical reasoning with incomplete commonsense domain knowledge, and incremental inductive learning, to guide the construction of deep network models from a small number of training examples. Experimental results in the context of a robot reasoning about the partial occlusion of objects and the stability of object configurations in simulated images indicate an improvement in reliability and a reduction in computational effort in comparison with an architecture based just on deep networks.

Recognizing, reasoning about, and providing understandable descriptions of spatial relations between objects is an important task for robots interacting with humans. This paper describes an architecture for incrementally learning and revising the grounding of spatial relations between objects. Answer Set Prolog, a declarative language, is used to represent and reason with incomplete knowledge that includes prepositional spatial relations between scene objects. A generic grounding of prepositions for spatial relations, human input (when available), and non-monotonic logical inference, are used to infer spatial relations between 3D point clouds in given scenes, incrementally acquiring a specialized metric grounding of the prepositions and the relative confidence associated with each grounding. The architecture is evaluated on a benchmark dataset of tabletop images and on complex simulated scenes of furniture.