Ugo Fiore

Federated multi-scale contrastive learning for semi-supervised anomaly detection in multimodal industrial Internet of Things

HFMD-Net: focal modulation and deformable convolutions for explainable pancreatic tumor segmentation with global–local boundary refinement

F focal-modulation-driven global context with deformable-convolution-based local adaptation produces boundary-refined pancreatic tumor segmentations on CT, which improves boundary fidelity and provides a transparent layer of review that may support treatment planning and longitudinal monitoring.

Solar Energy Production Forecasting With Hyperparameter-Optimized Time-Series Dense Encoder

This work introduces a fully automatic photovoltaic forecasting framework that combines the time-series dense encoder architecture with a genetic hyperparameter space explorer based on the non-dominated sorting genetic algorithm III to provide a promising approach for solar power forecasting.