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 5 Different frameworks’ UAR result for classification subtasks of three 3-class paralinguistic singing attributes

	Unweighted average recall (UAR) [%]
Frameworks	Front placement singing	Back placement singing	Vibrato	Average
SF-SVM (ComparE09)	31.87	34.91	35.52	34.10
SF-SVM (ComparE016)	33.7	33.76	42.84	36.77
E2E (ResNet)	35.2	36.2	41.89	37.76
E2E (Transformer)	33.6	39.42	38.97	37.33
DE-SVM (ResNet)	33.22	33.76	47.02	38
DE-SVM (Transformer)	30.61	36.71	43.67	37