Interpretable visual feature discovery using multiple instance learning: a case study of colonial Korean print