Automatic segmentation of ethnomusicological field recordings

Matija Marolt (Avtor), Ciril Bohak (Avtor), Alenka Kavčič (Avtor), Matevž Pesek (Avtor)

Povzetek

The article presents a method for segmentation of ethnomusicological field recordings. Field recordings are integral documents of folk music performances captured in the field, and typically contain performances, intertwined with interviews and commentaries. As these are live recordings, captured in non-ideal conditions, they usually contain significant background noise. We present a segmentation method that segments field recordings into individual units labelled as speech, solo singing, choir singing, and instrumentals. Classification is based on convolutional deep networks, and is augmented with a probabilistic approach for segmentation. We describe the dataset gathered for the task and the tools developed for gathering the reference annotations. We outline a deep network architecture based on residual modules for labelling short audio segments and compare it to the more standard feature based approaches, where an improvement in classification accuracy of over 10% was obtained. We also present the SeFiRe segmentation tool that incorporates the presented segmentation method.

Ključne besede

segmentacija zvočnih posnetkov;terenski posnetki;globoko učenje;pridobivanje informacij iz glasbe;audio segmentation;field recordings;deep learning;music information retrieval;

Podatki

Jezik:	Angleški jezik
Leto izida:	2019
Tipologija:	1.01 - Izvirni znanstveni članek
Organizacija:	UL FRI - Fakulteta za računalništvo in informatiko
UDK:	004:78
COBISS:	1538109123
ISSN:	2076-3417
Št. ogledov:	200
Št. prenosov:	56
Ocena:	0 (0 glasov)
Metapodatki:

Ostali podatki

Sekundarni jezik:	Slovenski jezik
Sekundarne ključne besede:	segmentacija zvočnih posnetkov;terenski posnetki;globoko učenje;pridobivanje informacij iz glasbe;
Vrsta dela (COBISS):	Članek v reviji
Strani:	str. 1-12
Letnik:	ǂVol. ǂ9
Zvezek:	ǂiss. ǂ3
Čas izdaje:	Jan. 2019
DOI:	10.3390/app9030439
ID:	13668074