Towards Explainable Neural-Symbolic Visual Reasoning

TL;DR

This paper advocates for neural-symbolic methods to enhance explainability in visual reasoning, proposing a model that explains neural decisions and can correct biases, demonstrated through an image captioning example.

Contribution

It introduces a reasoning model combining neural and symbolic approaches to generate explanations and correct biases in visual reasoning tasks.

Findings

Proposes a neural-symbolic explanation model

Demonstrates bias correction in neural decision rationale

Provides an example in image captioning context

Abstract

Many high-performance models suffer from a lack of interpretability. There has been an increasing influx of work on explainable artificial intelligence (XAI) in order to disentangle what is meant and expected by XAI. Nevertheless, there is no general consensus on how to produce and judge explanations. In this paper, we discuss why techniques integrating connectionist and symbolic paradigms are the most efficient solutions to produce explanations for non-technical users and we propose a reasoning model, based on definitions by Doran et al. [2017] (arXiv:1710.00794) to explain a neural network's decision. We use this explanation in order to correct bias in the network's decision rationale. We accompany this model with an example of its potential use, based on the image captioning method in Burns et al. [2018] (arXiv:1803.09797).