Curse to Blessing: Leveraging Model Inversion Attacks to Detect Backdoor Attacks in Federated Learning | Publicación