FLIP: a tool for explainable federated learning for image processing

In recent years, the adoption of Artificial Intelligence has grown significantly, driving the development of innovative methodologies to enhance AI-based models, such as Federated Machine Learning, an approach where multiple devices collaboratively train a shared model while keeping raw data localized, ensuring that only model updates are transmitted to a central server. In this paper, we present FLIP, a novel framework for image processing that leverages the principles of Federated Machine Learning. The framework supports key phases of federated training and incorporates explainability, with the aim to provide insights behind model predictions and improve interpretability.