1 paper
Francesco Di Salvo, David Tafler, Sebastian Doerrich +1
Large and well-annotated datasets are essential for advancing deep learning applications, however often costly or impossible to obtain by a single entity. In many areas, including…