Supervised Hierarchical Segmentation for Bird Bioacoustics Public Deposited

http://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/f1881q60d

Descriptions

Attribute NameValues
Creator
Abstract or Summary
  • Bioacoustics analysis can be used to conduct environmental monitoring by detecting the presence of birds species. This analysis usually involves identifying the species from their calls. In most frameworks, bird song syllables are extracted from audio recordings and individual syllables are input to a classifier to identify the species. Extraction of bird song syllables from audio recordings involves segmenting the bird song signal into individual syllables. However, syllable extraction from in-field recordings poses a challenge due to the presence of environmental noise. For such noisy recordings, supervised segmentation has been observed to perform better than unsupervised approaches. To perform segmentation, recordings are commonly converted to a time-frequency spectrogram. Supervision can then be provided at pixel level and syllable level. In pixel-level supervision, individual pixels are predicted to belong to a syllable, while in syllable-level supervision, the prediction is made for groups of pixels. In this thesis, we propose a supervised hierarchical segmentation approach that learns from both pixel and syllable levels supervision. Experimental results show that the proposed method outperforms existing supervised method that learns only at the pixel level.
Resource Type
Date Available
Date Copyright
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Non-Academic Affiliation
Keyword
Subject
Rights Statement
Peer Reviewed
Language
Replaces
Additional Information
  • description.provenance : Approved for entry into archive by Laura Wilson(laura.wilson@oregonstate.edu) on 2015-06-09T16:05:54Z (GMT) No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: bb87e2fb4674c76d0d2e9ed07fbb9c86 (MD5) TjahjaTeresaV2015.pdf: 506970 bytes, checksum: 00ec6c089efda78728f2885f1bb428e6 (MD5)
  • description.provenance : Submitted by Teresa Tjahja (tjahjat@onid.orst.edu) on 2015-06-05T22:11:59Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: bb87e2fb4674c76d0d2e9ed07fbb9c86 (MD5) TjahjaTeresaV2015.pdf: 506970 bytes, checksum: 00ec6c089efda78728f2885f1bb428e6 (MD5)
  • description.provenance : Approved for entry into archive by Julie Kurtz(julie.kurtz@oregonstate.edu) on 2015-06-08T16:55:57Z (GMT) No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: bb87e2fb4674c76d0d2e9ed07fbb9c86 (MD5) TjahjaTeresaV2015.pdf: 506970 bytes, checksum: 00ec6c089efda78728f2885f1bb428e6 (MD5)
  • description.provenance : Made available in DSpace on 2015-06-09T16:05:54Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: bb87e2fb4674c76d0d2e9ed07fbb9c86 (MD5) TjahjaTeresaV2015.pdf: 506970 bytes, checksum: 00ec6c089efda78728f2885f1bb428e6 (MD5) Previous issue date: 2015-06-01

Relationships

Parents:

This work has no parents.

Last modified

Downloadable Content

Download PDF

Items