Interactive Explainable Anomaly Detection for Industrial Settings

TL;DR

This paper introduces NearCAIPI, an interactive framework for explainable anomaly detection in industrial RGB images, combining CNN-based models, model-agnostic explanations, and user feedback to enhance trust and accuracy.

Contribution

It develops an interactive explanation framework that integrates human feedback into CNN-based anomaly detection, improving interpretability and trustworthiness in industrial settings.

Findings

Enhanced model trust through user interaction

Effective integration of human feedback into anomaly detection

Improved confidence in model decisions

Abstract

Being able to recognise defects in industrial objects is a key element of quality assurance in production lines. Our research focuses on visual anomaly detection in RGB images. Although Convolutional Neural Networks (CNNs) achieve high accuracies in this task, end users in industrial environments receive the model's decisions without additional explanations. Therefore, it is of interest to enrich the model's outputs with further explanations to increase confidence in the model and speed up anomaly detection. In our work, we focus on (1) CNN-based classification models and (2) the further development of a model-agnostic explanation algorithm for black-box classifiers. Additionally, (3) we demonstrate how we can establish an interactive interface that allows users to further correct the model's output. We present our NearCAIPI Interaction Framework, which improves AI through user interaction, and show how this approach increases the system's trustworthiness. We also illustrate how NearCAIPI can integrate human feedback into an interactive process chain.