pix2rule: End-to-end Neuro-symbolic Rule Learning
This addresses the challenge of integrating low-level visual input with high-level symbolic reasoning for AI systems, representing a novel approach rather than an incremental improvement.
The paper tackles the problem of combining visual processing with symbolic reasoning by introducing an end-to-end neuro-symbolic method that extracts symbolic relations and rules from images. It demonstrates that this model scales beyond state-of-the-art symbolic learners and outperforms deep relational neural networks on tasks like subgraph isomorphism and image classification with compound relations.
Humans have the ability to seamlessly combine low-level visual input with high-level symbolic reasoning often in the form of recognising objects, learning relations between them and applying rules. Neuro-symbolic systems aim to bring a unifying approach to connectionist and logic-based principles for visual processing and abstract reasoning respectively. This paper presents a complete neuro-symbolic method for processing images into objects, learning relations and logical rules in an end-to-end fashion. The main contribution is a differentiable layer in a deep learning architecture from which symbolic relations and rules can be extracted by pruning and thresholding. We evaluate our model using two datasets: subgraph isomorphism task for symbolic rule learning and an image classification domain with compound relations for learning objects, relations and rules. We demonstrate that our model scales beyond state-of-the-art symbolic learners and outperforms deep relational neural network architectures.