PocketEAR: An Assistive Sound Classification System for Hearing-Impaired

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F19%3A43955841" target="_blank" >RIV/49777513:23520/19:43955841 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1007/978-3-030-26061-3" target="_blank" >http://dx.doi.org/10.1007/978-3-030-26061-3</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/978-3-030-26061-3" target="_blank" >10.1007/978-3-030-26061-3</a>

Jazyk výsledku

angličtina

Název v původním jazyce

PocketEAR: An Assistive Sound Classification System for Hearing-Impaired

Popis výsledku v původním jazyce

This paper describes the design and operation of an assistive system called PocketEAR which is primarily targeted towards hearing-impaired users. It helps them with orientation in acoustically active environments by continuously monitoring and classifying the incoming sounds and displaying the captured sound classes to the users. The environmental sound recognizer is designed as a two-stage deep convolutional neural network classifier (consists of the so-called superclassifier and a set of the so-called subclassifiers) fed with sequences of MFCC vectors. It is wrapped in a distributed client-server system where the sound capturing in terrain, (pre)processing and displaying of the classication results are performed by instances of a mobile client application, and the actual classication and maintenance are carried out by two co-operating servers. The paper discusses in details the architecture of the environmental sound classier as well as the used task-specific sound processing

Název v anglickém jazyce

PocketEAR: An Assistive Sound Classification System for Hearing-Impaired

Popis výsledku anglicky

This paper describes the design and operation of an assistive system called PocketEAR which is primarily targeted towards hearing-impaired users. It helps them with orientation in acoustically active environments by continuously monitoring and classifying the incoming sounds and displaying the captured sound classes to the users. The environmental sound recognizer is designed as a two-stage deep convolutional neural network classifier (consists of the so-called superclassifier and a set of the so-called subclassifiers) fed with sequences of MFCC vectors. It is wrapped in a distributed client-server system where the sound capturing in terrain, (pre)processing and displaying of the classication results are performed by instances of a mobile client application, and the actual classication and maintenance are carried out by two co-operating servers. The paper discusses in details the architecture of the environmental sound classier as well as the used task-specific sound processing

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2019

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Speech and Computer

ISBN

978-3-030-26060-6

ISSN

0302-9743

e-ISSN

—

Počet stran výsledku

11

Strana od-do

82-92

Název nakladatele

Springer

Místo vydání

Cham

Místo konání akce

Istanbul

Datum konání akce

20. 8. 2019

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—