Sign Language Numeral Gestures Recognition Using Convolutional Neural Network
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F18%3A43952760" target="_blank" >RIV/49777513:23520/18:43952760 - isvavai.cz</a>
Result on the web
<a href="https://link.springer.com/chapter/10.1007%2F978-3-319-99582-3_8" target="_blank" >https://link.springer.com/chapter/10.1007%2F978-3-319-99582-3_8</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1007/978-3-319-99582-3_8" target="_blank" >10.1007/978-3-319-99582-3_8</a>
Alternative languages
Result language
angličtina
Original language name
Sign Language Numeral Gestures Recognition Using Convolutional Neural Network
Original language description
This paper presents usage of convolutional neural network for classification of sign language numeral gestures. For requirements of this research, we created a new dataset of these gestures. The dataset was recorded via Kinect v2 device and it consists of recordings of 18 different people. Only depth data-stream was used in our research. For a classification task, there was utilized classic VGG16 architecture and its results were compared with chosen baseline method and other tested architectures. Our experiment on classification showed the great potential of neural networks for this task. We reached recognition accuracy 86.45%, which is by more than 34% better result than chosen baseline method.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
20205 - Automation and control systems
Result continuities
Project
—
Continuities
S - Specificky vyzkum na vysokych skolach
Others
Publication year
2018
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Interactive Collaborative Robotics Third International Conference, ICR 2018 Leipzig, Germany, September 18–22, 2018 Proceedings
ISBN
978-3-319-99581-6
ISSN
0302-9743
e-ISSN
1611-3349
Number of pages
8
Pages from-to
70-77
Publisher name
Springer Nature Switzerland AG
Place of publication
Cham
Event location
Leipzig, Germany
Event date
Sep 18, 2018
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—