Zobrazit v EDS

Solar-Powered Deep Learning-Based Recognition System of Daily Used Objects and Human Faces for Assistance of the Visually Impaired

Uloženo v:
Podrobná bibliografie
Název: Solar-Powered Deep Learning-Based Recognition System of Daily Used Objects and Human Faces for Assistance of the Visually Impaired
Autoři: Bernardo Calabrese, Ramiro Velázquez, Carolina Del-Valle-Soto, Roberto de Fazio, Nicola Ivan Giannoccaro, Paolo Visconti
Zdroj: Energies, Vol 13, Iss 22, p 6104 (2020)
Informace o vydavateli: MDPI AG, 2020.
Rok vydání: 2020
Sbírka: LCC:Technology
Témata: assistive technology, convolutional neural networks (CNN), deep learning, faster R-CNN, mobile computing, object recognition, Technology
Popis: This paper introduces a novel low-cost solar-powered wearable assistive technology (AT) device, whose aim is to provide continuous, real-time object recognition to ease the finding of the objects for visually impaired (VI) people in daily life. The system consists of three major components: a miniature low-cost camera, a system on module (SoM) computing unit, and an ultrasonic sensor. The first is worn on the user’s eyeglasses and acquires real-time video of the nearby space. The second is worn as a belt and runs deep learning-based methods and spatial algorithms which process the video coming from the camera performing objects’ detection and recognition. The third assists on positioning the objects found in the surrounding space. The developed device provides audible descriptive sentences as feedback to the user involving the objects recognized and their position referenced to the user gaze. After a proper power consumption analysis, a wearable solar harvesting system, integrated with the developed AT device, has been designed and tested to extend the energy autonomy in the different operating modes and scenarios. Experimental results obtained with the developed low-cost AT device have demonstrated an accurate and reliable real-time object identification with an 86% correct recognition rate and 215 ms average time interval (in case of high-speed SoM operating mode) for the image processing. The proposed system is capable of recognizing the 91 objects offered by the Microsoft Common Objects in Context (COCO) dataset plus several custom objects and human faces. In addition, a simple and scalable methodology for using image datasets and training of Convolutional Neural Networks (CNNs) is introduced to add objects to the system and increase its repertory. It is also demonstrated that comprehensive trainings involving 100 images per targeted object achieve 89% recognition rates, while fast trainings with only 12 images achieve acceptable recognition rates of 55%.
Druh dokumentu: article
Popis souboru: electronic resource
Jazyk: English
ISSN: 1996-1073
Relation: https://www.mdpi.com/1996-1073/13/22/6104; https://doaj.org/toc/1996-1073
DOI: 10.3390/en13226104
Přístupová URL adresa: https://doaj.org/article/4780c997e3b84e5da24de71cc0659b9d
Přístupové číslo: edsdoj.4780c997e3b84e5da24de71cc0659b9d
Databáze: Directory of Open Access Journals
FullText Text:
  Availability: 0
CustomLinks:
  – Url: https://doaj.org/article/4780c997e3b84e5da24de71cc0659b9d
    Name: EDS - DOAJ (s4221598)
    Category: fullText
    Text: View record in DOAJ
  – Url: https://resolver.ebscohost.com/openurl?sid=EBSCO:edsdoj&genre=article&issn=19961073&ISBN=&volume=13&issue=22&date=20201101&spage=6104&pages=6104-6104&title=Energies&atitle=Solar-Powered%20Deep%20Learning-Based%20Recognition%20System%20of%20Daily%20Used%20Objects%20and%20Human%20Faces%20for%20Assistance%20of%20the%20Visually%20Impaired&aulast=Bernardo%20Calabrese&id=DOI:10.3390/en13226104
    Name: Full Text Finder
    Category: fullText
    Text: Full Text Finder
    Icon: https://imageserver.ebscohost.com/branding/images/FTF.gif
    MouseOverText: Full Text Finder
  – Url: https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=EBSCO&SrcAuth=EBSCO&DestApp=WOS&ServiceName=TransferToWoS&DestLinkType=GeneralSearchSummary&Func=Links&author=Calabrese%20B
    Name: ISI
    Category: fullText
    Text: Nájsť tento článok vo Web of Science
    Icon: https://imagesrvr.epnet.com/ls/20docs.gif
    MouseOverText: Nájsť tento článok vo Web of Science
Header DbId: edsdoj
DbLabel: Directory of Open Access Journals
An: edsdoj.4780c997e3b84e5da24de71cc0659b9d
RelevancyScore: 880
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 880.150939941406
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Solar-Powered Deep Learning-Based Recognition System of Daily Used Objects and Human Faces for Assistance of the Visually Impaired
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Bernardo+Calabrese%22">Bernardo Calabrese</searchLink><br /><searchLink fieldCode="AR" term="%22Ramiro+Velázquez%22">Ramiro Velázquez</searchLink><br /><searchLink fieldCode="AR" term="%22Carolina+Del-Valle-Soto%22">Carolina Del-Valle-Soto</searchLink><br /><searchLink fieldCode="AR" term="%22Roberto+de+Fazio%22">Roberto de Fazio</searchLink><br /><searchLink fieldCode="AR" term="%22Nicola+Ivan+Giannoccaro%22">Nicola Ivan Giannoccaro</searchLink><br /><searchLink fieldCode="AR" term="%22Paolo+Visconti%22">Paolo Visconti</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: Energies, Vol 13, Iss 22, p 6104 (2020)
– Name: Publisher
  Label: Publisher Information
  Group: PubInfo
  Data: MDPI AG, 2020.
– Name: DatePubCY
  Label: Publication Year
  Group: Date
  Data: 2020
– Name: Subset
  Label: Collection
  Group: HoldingsInfo
  Data: LCC:Technology
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22assistive+technology%22">assistive technology</searchLink><br /><searchLink fieldCode="DE" term="%22convolutional+neural+networks+%28CNN%29%22">convolutional neural networks (CNN)</searchLink><br /><searchLink fieldCode="DE" term="%22deep+learning%22">deep learning</searchLink><br /><searchLink fieldCode="DE" term="%22faster+R-CNN%22">faster R-CNN</searchLink><br /><searchLink fieldCode="DE" term="%22mobile+computing%22">mobile computing</searchLink><br /><searchLink fieldCode="DE" term="%22object+recognition%22">object recognition</searchLink><br /><searchLink fieldCode="DE" term="%22Technology%22">Technology</searchLink>
– Name: Abstract
  Label: Description
  Group: Ab
  Data: This paper introduces a novel low-cost solar-powered wearable assistive technology (AT) device, whose aim is to provide continuous, real-time object recognition to ease the finding of the objects for visually impaired (VI) people in daily life. The system consists of three major components: a miniature low-cost camera, a system on module (SoM) computing unit, and an ultrasonic sensor. The first is worn on the user’s eyeglasses and acquires real-time video of the nearby space. The second is worn as a belt and runs deep learning-based methods and spatial algorithms which process the video coming from the camera performing objects’ detection and recognition. The third assists on positioning the objects found in the surrounding space. The developed device provides audible descriptive sentences as feedback to the user involving the objects recognized and their position referenced to the user gaze. After a proper power consumption analysis, a wearable solar harvesting system, integrated with the developed AT device, has been designed and tested to extend the energy autonomy in the different operating modes and scenarios. Experimental results obtained with the developed low-cost AT device have demonstrated an accurate and reliable real-time object identification with an 86% correct recognition rate and 215 ms average time interval (in case of high-speed SoM operating mode) for the image processing. The proposed system is capable of recognizing the 91 objects offered by the Microsoft Common Objects in Context (COCO) dataset plus several custom objects and human faces. In addition, a simple and scalable methodology for using image datasets and training of Convolutional Neural Networks (CNNs) is introduced to add objects to the system and increase its repertory. It is also demonstrated that comprehensive trainings involving 100 images per targeted object achieve 89% recognition rates, while fast trainings with only 12 images achieve acceptable recognition rates of 55%.
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: article
– Name: Format
  Label: File Description
  Group: SrcInfo
  Data: electronic resource
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1996-1073
– Name: NoteTitleSource
  Label: Relation
  Group: SrcInfo
  Data: https://www.mdpi.com/1996-1073/13/22/6104; https://doaj.org/toc/1996-1073
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.3390/en13226104
– Name: URL
  Label: Access URL
  Group: URL
  Data: <link linkTarget="URL" linkTerm="https://doaj.org/article/4780c997e3b84e5da24de71cc0659b9d" linkWindow="_blank">https://doaj.org/article/4780c997e3b84e5da24de71cc0659b9d</link>
– Name: AN
  Label: Accession Number
  Group: ID
  Data: edsdoj.4780c997e3b84e5da24de71cc0659b9d
PLink https://erproxy.cvtisr.sk/sfx/access?url=https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsdoj&AN=edsdoj.4780c997e3b84e5da24de71cc0659b9d
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.3390/en13226104
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 1
        StartPage: 6104
    Subjects:
      – SubjectFull: assistive technology
        Type: general
      – SubjectFull: convolutional neural networks (CNN)
        Type: general
      – SubjectFull: deep learning
        Type: general
      – SubjectFull: faster R-CNN
        Type: general
      – SubjectFull: mobile computing
        Type: general
      – SubjectFull: object recognition
        Type: general
      – SubjectFull: Technology
        Type: general
    Titles:
      – TitleFull: Solar-Powered Deep Learning-Based Recognition System of Daily Used Objects and Human Faces for Assistance of the Visually Impaired
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Bernardo Calabrese
      – PersonEntity:
          Name:
            NameFull: Ramiro Velázquez
      – PersonEntity:
          Name:
            NameFull: Carolina Del-Valle-Soto
      – PersonEntity:
          Name:
            NameFull: Roberto de Fazio
      – PersonEntity:
          Name:
            NameFull: Nicola Ivan Giannoccaro
      – PersonEntity:
          Name:
            NameFull: Paolo Visconti
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 11
              Type: published
              Y: 2020
          Identifiers:
            – Type: issn-print
              Value: 19961073
          Numbering:
            – Type: volume
              Value: 13
            – Type: issue
              Value: 22
          Titles:
            – TitleFull: Energies
              Type: main
ResultId 1