IMPACTX: improving model performance by appropriately constraining the training with teacher explanations

TL;DR

IMPACTX introduces a novel XAI-based attention mechanism that automatically enhances deep learning model performance and provides explanations without external tools during inference.

Contribution

The paper presents IMPACTX, a new method that leverages XAI outputs as an automated attention mechanism to improve model accuracy and interpretability.

Findings

IMPACTX improves performance across multiple DL models and datasets.

It provides accurate feature attribution maps without external XAI during inference.

Consistent performance gains observed on CIFAR-10, CIFAR-100, and STL-10.

Abstract

The eXplainable Artificial Intelligence (XAI) research predominantly concentrates to provide explainations about AI model decisions, especially Deep Learning (DL) models. However, there is a growing interest in using XAI techniques to automatically improve the performance of the AI systems themselves. This paper proposes IMPACTX, a novel approach that leverages XAI as a fully automated attention mechanism, without requiring external knowledge or human feedback. Experimental results show that IMPACTX has improved performance respect to the standalone ML model by integrating an attention mechanism based an XAI method outputs during the model training. Furthermore, IMPACTX directly provides proper feature attribution maps for the model's decisions, without relying on external XAI methods during the inference process. Our proposal is evaluated using three widely recognized DL models (EfficientNet-B2, MobileNet, and LeNet-5) along with three standard image datasets: CIFAR-10, CIFAR-100, and STL-10. The results show that IMPACTX consistently improves the performance of all the inspected DL models across all evaluated datasets, and it directly provides appropriate explanations for its responses.