Conference paper Open Access

Short-term Recognition of Human Activities using Convolutional Neural Networks

M.Papakostas; T. Giannakopoulos; F. Makedon; V. Karkaletsis


MARC21 XML Export

<?xml version='1.0' encoding='UTF-8'?>
<record xmlns="http://www.loc.gov/MARC21/slim">
  <leader>00000nam##2200000uu#4500</leader>
  <controlfield tag="005">20170908072921.0</controlfield>
  <controlfield tag="001">376482</controlfield>
  <datafield tag="711" ind1=" " ind2=" ">
    <subfield code="g">SITIS</subfield>
    <subfield code="a">2016 International Conference on Signal-Image Technology &amp; Internet Based Systems, IEEE, 2016</subfield>
  </datafield>
  <datafield tag="700" ind1=" " ind2=" ">
    <subfield code="u">NCSRD</subfield>
    <subfield code="a">T. Giannakopoulos</subfield>
  </datafield>
  <datafield tag="700" ind1=" " ind2=" ">
    <subfield code="u">U Texas Arlington</subfield>
    <subfield code="a">F. Makedon</subfield>
  </datafield>
  <datafield tag="700" ind1=" " ind2=" ">
    <subfield code="u">NCSRD</subfield>
    <subfield code="a">V. Karkaletsis</subfield>
  </datafield>
  <datafield tag="856" ind1="4" ind2=" ">
    <subfield code="s">1589361</subfield>
    <subfield code="z">md5:04f985a6b7deee6ec6b131671261beb7</subfield>
    <subfield code="u">https://zenodo.org/record/376482/files/SITIS_MIRA_2016.pdf</subfield>
  </datafield>
  <datafield tag="542" ind1=" " ind2=" ">
    <subfield code="l">open</subfield>
  </datafield>
  <datafield tag="260" ind1=" " ind2=" ">
    <subfield code="c">2017-03-10</subfield>
  </datafield>
  <datafield tag="909" ind1="C" ind2="O">
    <subfield code="p">openaire</subfield>
    <subfield code="p">user-radio</subfield>
    <subfield code="o">oai:zenodo.org:376482</subfield>
  </datafield>
  <datafield tag="100" ind1=" " ind2=" ">
    <subfield code="u">NCSRD</subfield>
    <subfield code="a">M.Papakostas</subfield>
  </datafield>
  <datafield tag="245" ind1=" " ind2=" ">
    <subfield code="a">Short-term Recognition of Human Activities using Convolutional Neural Networks</subfield>
  </datafield>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">user-radio</subfield>
  </datafield>
  <datafield tag="536" ind1=" " ind2=" ">
    <subfield code="c">643892</subfield>
    <subfield code="a">Robots in assisted living environments: Unobtrusive, efficient, reliable and modular solutions for independent ageing</subfield>
  </datafield>
  <datafield tag="540" ind1=" " ind2=" ">
    <subfield code="u">http://creativecommons.org/licenses/by/4.0/legalcode</subfield>
    <subfield code="a">Creative Commons Attribution 4.0 International</subfield>
  </datafield>
  <datafield tag="650" ind1="1" ind2="7">
    <subfield code="a">cc-by</subfield>
    <subfield code="2">opendefinition.org</subfield>
  </datafield>
  <datafield tag="520" ind1=" " ind2=" ">
    <subfield code="a">&lt;p&gt;This paper proposes a deep learning classification method for frame-wise recognition of human activities, using raw color (RGB) information. In particular, we present a Convolutional Neural Network (CNN) classification approach for recognising three basic motion activity classes, that cover the vast majority of human activities in the context of a home monitoring environment, namely: sitting, walking and standing up. A real-world fully annotated dataset has been compiled, in the context of an assisted living home environment. Through extensive experimentation we have highlighted the benefits of deep learning architectures against traditional shallow classifiers functioning on hand-crafted features, on the task of activity recognition. Our approach proves the robustness and the quality of CNN classifiers that lies on learning highly invariant features. Our ultimate goal is to tackle the challenging task of activity recognition in environments that are characterized with high levels of inherent noise.&lt;/p&gt;</subfield>
  </datafield>
  <datafield tag="024" ind1=" " ind2=" ">
    <subfield code="a">10.5281/zenodo.376482</subfield>
    <subfield code="2">doi</subfield>
  </datafield>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">publication</subfield>
    <subfield code="b">conferencepaper</subfield>
  </datafield>
</record>
43
35
views
downloads
All versions This version
Views 4343
Downloads 3535
Data volume 55.6 MB55.6 MB
Unique views 4242
Unique downloads 3333

Share

Cite as