The issue of homogeneity in performance assessment of proposed algorithms for information extraction is generally perceived also in the Earth Observation (EO) domain. Different authors propose different datasets to test their developed algorithms and to the reader it is frequently difficult to assess which is better for his/her specific application, given the wide variability in test sets that makes pure comparison of e.g. accuracy values less meaningful than one would desire. With our work, we gave a modest contribution to ease the problem by making it possible to automatically distribute a limited set of possible "standard" open datasets, together with some ground truth info, and automatically assess processing results provided by the users.
File in questo prodotto:
Non ci sono file associati a questo prodotto.