Endre søk
RefereraExporteraLink to record
Permanent link

Direct link
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
From Weak to Strong Sound Event Labels using Adaptive Change-Point Detection and Active Learning
RISE Research Institutes of Sweden, Digitala system, Datavetenskap.ORCID-id: 0000-0002-5032-4367
RISE Research Institutes of Sweden, Digitala system, Datavetenskap.ORCID-id: 0000-0002-9567-2218
Lund University, Sweden.
Tampere University, Finland.
2024 (engelsk)Konferansepaper, Publicerat paper (Fagfellevurdert)
Abstract [en]

We propose an adaptive change point detection method (A-CPD) for machine guided weak label annotation of audio recording segments. The goal is to maximize the amount of information gained about the temporal activations of the target sounds. For each unlabeled audio recording, we use a prediction model to derive a probability curve used to guide annotation. The prediction model is initially pre-trained on available annotated sound event data with classes that are disjoint from the classes in the unlabeled dataset. The prediction model then gradually adapts to the annotations provided by the annotator in an active learning loop. We derive query segments to guide the weak label annotator towards strong labels, using change point detection on these probabilities. We show that it is possible to derive strong labels of high quality with a limited annotation budget, and show favorable results for A-CPD when compared to two baseline query segment strategies. 

sted, utgiver, år, opplag, sider
European Signal Processing Conference, EUSIPCO , 2024. s. 902-906
Emneord [en]
Adversarial machine learning; Audio recordings; Budget control; Change detection; Contrastive Learning; Deep learning; Prediction models; Sound recording; Active Learning; Annotation; Change point detection; Deep learning; Detection methods; Prediction modelling; Query segments; Sound event detection; Sound events; Weak labels; Active learning
HSV kategori
Identifikatorer
URN: urn:nbn:se:ri:diva-76157Scopus ID: 2-s2.0-85208422384ISBN: 9789464593617 (digital)OAI: oai:DiVA.org:ri-76157DiVA, id: diva2:1914551
Konferanse
32nd European Signal Processing Conference, EUSIPCO 2024. Lyon. 26 August 2024 through 30 August 2024
Merknad

This work was supported by The Swedish Foundation for Strategic Research (SSF; FID20-0028) and Sweden\u2019s Innovation Agency (2023-01486).

Tilgjengelig fra: 2024-11-19 Laget: 2024-11-19 Sist oppdatert: 2024-11-19bibliografisk kontrollert

Open Access i DiVA

Fulltekst mangler i DiVA

Scopus

Person

Martinsson, JohnMogren, Olof

Søk i DiVA

Av forfatter/redaktør
Martinsson, JohnMogren, Olof
Av organisasjonen

Søk utenfor DiVA

GoogleGoogle Scholar

isbn
urn-nbn

Altmetric

isbn
urn-nbn
Totalt: 11 treff
RefereraExporteraLink to record
Permanent link

Direct link
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
v. 2.45.0