1 paper · 1 filter
Fabio De Gaspari, Dorjan Hitaj, Luigi V. Mancini
The unprecedented availability of training data fueled the rapid development of powerful neural networks in recent years. However, the need for such large amounts of data leads to…