Paralinguistic singing attribute recognition using supervised machine learning for describing the classical tenor solo singing voice in vocal pedagogy

EURASIP Journal on Audio, Speech, and Music Processing

Table 6 Different frameworks’ UAR result for binary classification subtask of roughness

	Unweighted average recall (UAR) [%]
Frameworks	Roughness
SF-SVM (ComparE09)	51.85
SF-SVM (ComparE016)	55.19
E2E (ResNet)	56.23
E2E (Transformer)	55.39
DE-SVM (ResNet)	58.83
DE-SVM (Transformer)	54.4