Information Management and Big Data by Unknown

Author:Unknown
Language: eng
Format: epub
ISBN: 9783030461409
Publisher: Springer International Publishing

1 Introduction

Currently, people don’t have the need to learn SL that is unique to each community and designed by themselves, i.e. the spanish SL is not completely similar as peruvian although them speak spanish. There are models that use CNN, Recurrent Neural Network (RNN) and Bidirectional Recurrent Neural Network (BRNN) that get good results for their SLs; however, they lose useful information by capturing only one type of data. Create a specific dataset for a SL is complicated because they have their own large dictionary; therefore, it is not possible to apply the same models and dataset for peruvian context. For developing an architecture that achieves an acceptable success rate, it is necessary to identify techniques as DNN (Deep Neural Network) and Natural Language Processing (NLP), for build our model with CNN along with RNN we used techniques as early stoping (ES), transfer learning (TL) and dropout; in addition to, our dictionary is done with expressions rather than just using simple words, to have better accuracy we use three types of stream: rgb, detph and skeleton which provide more signal information regarding distance and position of the gesture.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.