Title: | Framework for automated customer service in sign language |
Authors: | Malawski, Filip Gałka, Jakub |
Citation: | WSCG '2016: short communications proceedings: The 24th International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision 2016 in co-operation with EUROGRAPHICS: University of West Bohemia, Plzen, Czech RepublicMay 30 - June 3 2016, p. 181-188. |
Issue Date: | 2016 |
Publisher: | Václav Skala - UNION Agency |
Document type: | konferenční příspěvek conferenceObject |
URI: | wscg.zcu.cz/WSCG2016/!!_CSRN-2602.pdf http://hdl.handle.net/11025/29702 |
ISBN: | 978-80-86943-58-9 |
ISSN: | 2464-4617 |
Keywords: | rozpoznávání znakového jazyka;Kinect 2;HMM;HCI;rámec |
Keywords in different language: | sign language recognition;Kinect 2;HMM;HCI;framework |
Abstract: | Deaf people need the help of an interpreter in formal relations, such as visiting offices or medical institutions. We present a new framework for building systems for sign language interaction, which can provide basic automated customer service for the deaf. The framework covers all steps required to build such a system from scratch - the acquisition of scenario-specific corpora, extraction of features, training of models, recognition, user interface, integration and configuration of the final application. The usability of the framework has been evaluated by creating a proof-of-concept system for automated scheduling of doctor's appointments in sign language. The results indicate that the process of building a sign language interaction system with our framework is relatively quick and simple. Recognition efficiency was evaluated as well and proved to be sufficient for practical use. |
Rights: | © Václav Skala - UNION Agency |
Appears in Collections: | WSCG '2016: Short Papers Proceedings |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
Malawski.pdf | Plný text | 1,02 MB | Adobe PDF | View/Open |
Please use this identifier to cite or link to this item:
http://hdl.handle.net/11025/29702
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.