Skip to main content

Table 4 NMF baseline classification accuracy (%)

From: High level feature extraction for the self-taught learning algorithm

Training Dictionary size K
Set 100 200 300 500
IS-20 52.7 53.1 54.2 54.5
IS-50 54.4 55.8 56.3 57.2
IS-100 52.7 54.0 55.3 56.9
IS-250 54.2 55.8 56.6 57.5
  1. Bases are learned from the labeled ISMIR training data.