Publikation:

Computer Vision and Deep Learning Methods for Measuring and Modeling Animal Behavior

Lade...
Vorschaubild

Dateien

Graving_2-dgcbudqch6ix8.pdf
Graving_2-dgcbudqch6ix8.pdfGröße: 34.14 MBDownloads: 548

Datum

2021

Herausgeber:innen

Kontakt

ISSN der Zeitschrift

Electronic ISSN

ISBN

Bibliografische Daten

Verlag

Schriftenreihe

Auflagebezeichnung

DOI (zitierfähiger Link)
ArXiv-ID

Internationale Patentnummer

Link zur Lizenz

Angaben zur Forschungsförderung

Projekt

Open Access-Veröffentlichung
Open Access Green
Core Facility der Universität Konstanz

Gesperrt bis

Titel in einer weiteren Sprache

Publikationstyp
Dissertation
Publikationsstatus
Published

Erschienen in

Zusammenfassung

The study of animal behavior is a fundamental pursuit for answering scientific questions across a variety of fields — including neuroscience, psychology, ecology, genetics, and evolution.While the task of collecting accurate and complete behavioral data has typically always been difficult, laborious, and subjective, in recent years there has been rapid progress in methods for automatically quantifying behavior objectively and at scale. This progress has been primarily driven by the emergence of new computational hardware, software, and algorithms for measuring behavior. In order to reveal core insights about how animals organize their behavior with the increased quality and resolution of these data comes the need for new methods for data-driven modeling. Here, in this thesis, I focus on computational tools—the development of new algorithms and software—for measuring and modeling behavior using methods from computer vision, deep learning, Bayesian inference, and probabilistic programming, while also synthesizing these approaches with ideas from other relevant areas such as information theory, nonlinear dynamics, and statistical physics. First, I developed a barcode tracking system for automated behavioral studies where the location and identity of individuals can be reliably tracked over several weeks (or potentially longer) using conventional computer vision (Chapter 1). Next, I developed general-purpose deep learning methods for measuring animal posture — any set of user-selected body parts — in the laboratory and field (Chapter 2). Finally, I introduce methods for using these posture data to model behavior with techniques from machine learning andBayesian statistical inference (Chapter 3). Together these methods reduce barriers to measuring and modeling animal behavior and allow researchers to answer scientific questions that were previously intractable

Zusammenfassung in einer weiteren Sprache

Fachgebiet (DDC)
570 Biowissenschaften, Biologie

Schlagwörter

Konferenz

Rezension
undefined / . - undefined, undefined

Forschungsvorhaben

Organisationseinheiten

Zeitschriftenheft

Zugehörige Datensätze in KOPS

Zitieren

ISO 690GRAVING, Jacob M., 2021. Computer Vision and Deep Learning Methods for Measuring and Modeling Animal Behavior [Dissertation]. Konstanz: University of Konstanz
BibTex
@phdthesis{Graving2021Compu-53120,
  year={2021},
  title={Computer Vision and Deep Learning Methods for Measuring and Modeling Animal Behavior},
  author={Graving, Jacob M.},
  address={Konstanz},
  school={Universität Konstanz}
}
RDF
<rdf:RDF
    xmlns:dcterms="http://purl.org/dc/terms/"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:bibo="http://purl.org/ontology/bibo/"
    xmlns:dspace="http://digital-repositories.org/ontologies/dspace/0.1.0#"
    xmlns:foaf="http://xmlns.com/foaf/0.1/"
    xmlns:void="http://rdfs.org/ns/void#"
    xmlns:xsd="http://www.w3.org/2001/XMLSchema#" > 
  <rdf:Description rdf:about="https://kops.uni-konstanz.de/server/rdf/resource/123456789/53120">
    <dcterms:available rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2021-03-10T09:45:47Z</dcterms:available>
    <dcterms:isPartOf rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/28"/>
    <dcterms:abstract xml:lang="eng">The study of animal behavior is a fundamental pursuit for answering scientific questions across a variety of fields — including neuroscience, psychology, ecology, genetics, and evolution.While the task of collecting accurate and complete behavioral data has typically always been difficult, laborious, and subjective, in recent years there has been rapid progress in methods for automatically quantifying behavior objectively and at scale. This progress has been primarily driven by the emergence of new computational hardware, software, and algorithms for measuring behavior. In order to reveal core insights about how animals organize their behavior with the increased quality and resolution of these data comes the need for new methods for data-driven modeling. Here, in this thesis, I focus on computational tools—the development of new algorithms and software—for measuring and modeling behavior using methods from computer vision, deep learning, Bayesian inference, and probabilistic programming, while also synthesizing these approaches with ideas from other relevant areas such as information theory, nonlinear dynamics, and statistical physics. First, I developed a barcode tracking system for automated behavioral studies where the location and identity of individuals can be reliably tracked over several weeks (or potentially longer) using conventional computer vision (Chapter 1). Next, I developed general-purpose deep learning methods for measuring animal posture — any set of user-selected body parts — in the laboratory and field (Chapter 2). Finally, I introduce methods for using these posture data to model behavior with techniques from machine learning andBayesian statistical inference (Chapter 3). Together these methods reduce barriers to measuring and modeling animal behavior and allow researchers to answer scientific questions that were previously intractable</dcterms:abstract>
    <dcterms:rights rdf:resource="http://creativecommons.org/licenses/by/4.0/"/>
    <dspace:isPartOfCollection rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/28"/>
    <dc:creator>Graving, Jacob M.</dc:creator>
    <dcterms:hasPart rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/53120/3/Graving_2-dgcbudqch6ix8.pdf"/>
    <dspace:hasBitstream rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/53120/3/Graving_2-dgcbudqch6ix8.pdf"/>
    <dcterms:issued>2021</dcterms:issued>
    <void:sparqlEndpoint rdf:resource="http://localhost/fuseki/dspace/sparql"/>
    <dc:contributor>Graving, Jacob M.</dc:contributor>
    <dc:date rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2021-03-10T09:45:47Z</dc:date>
    <bibo:uri rdf:resource="https://kops.uni-konstanz.de/handle/123456789/53120"/>
    <dc:rights>Attribution 4.0 International</dc:rights>
    <dc:language>eng</dc:language>
    <dcterms:title>Computer Vision and Deep Learning Methods for Measuring and Modeling Animal Behavior</dcterms:title>
    <foaf:homepage rdf:resource="http://localhost:8080/"/>
  </rdf:Description>
</rdf:RDF>

Interner Vermerk

xmlui.Submission.submit.DescribeStep.inputForms.label.kops_note_fromSubmitter

Kontakt
URL der Originalveröffentl.

Prüfdatum der URL

Prüfungsdatum der Dissertation

March 1, 2021
Hochschulschriftenvermerk
Konstanz, Univ., Diss., 2021
Finanzierungsart

Kommentar zur Publikation

Allianzlizenz
Corresponding Authors der Uni Konstanz vorhanden
Internationale Co-Autor:innen
Universitätsbibliographie
Begutachtet
Diese Publikation teilen