Object-Centric Neuro-Argumentative Learning

TL;DR

This paper presents a novel neural argumentative learning architecture that combines deep learning and symbolic reasoning for image analysis, enhancing interpretability and reliability.

Contribution

It introduces a hybrid neural-symbolic framework integrating object-centric learning with assumption-based argumentation for improved image-based decision making.

Findings

Competitive performance on synthetic datasets

Effective integration of neural and symbolic components

Enhanced interpretability of image analysis

Abstract

Over the last decade, as we rely more on deep learning technologies to make critical decisions, concerns regarding their safety, reliability and interpretability have emerged. We introduce a novel Neural Argumentative Learning (NAL) architecture that integrates Assumption-Based Argumentation (ABA) with deep learning for image analysis. Our architecture consists of neural and symbolic components. The former segments and encodes images into facts using object-centric learning, while the latter applies ABA learning to develop ABA frameworks enabling predictions with images. Experiments on synthetic data show that the NAL architecture can be competitive with a state-of-the-art alternative.