Xplique: A Deep Learning Explainability Toolbox

Fel, Thomas; Hervier, L.; Vigouroux, D.; Poche, A.; Plakoo, J.; Cadène, R.; Chalvidal, M.; Colin, J.; Boissin, T.; Bethune, L.; Picard, A.; Nicodeme, C.; Gardes, L.; Flandin, G.; Serre, T.

doi:https://doi.org/10.48550/arXiv.2206.04394

Xplique: A Deep Learning Explainability Toolbox

Authors: Fel, T., Hervier, L., Vigouroux, D., Poche, A., Plakoo, J., Cadène, R., Chalvidal, M., Colin, J. , Boissin, T., Bethune, L., Picard, A., Nicodeme, C., Gardes, L., Flandin, G., Serre, T.

External link: https://arxiv.org/abs/2206.04394
Publication: Conference on Computer Vision and Pattern Recognition (CVPR), Workshop on Explainable Artificial Intelligence for Computer Vision (XAI4CV) , 2022
DOI: https://doi.org/10.48550/arXiv.2206.04394
PDF: Click here for the PDF paper

Today’s most advanced machine-learning models are hardly scrutable. The key challenge for explainability methods is to help assisting researchers in opening up these black boxes –– by revealing the strategy that led to a given decision, by characterizing their internal states or by studying the underlying data representation. To address this challenge, we have developed Xplique: a software library for explainability which includes explainability methods as well as associated evaluation metrics. It interfaces with one of the most popular learning libraries: Tensorflow as well as other libraries including PyTorch, scikit-learn and Theano. The code is licensed under the MIT license and is freely available at github.com/deel-ai/xplique.