|
Title:
|
KNOWLEDGE BASE CREATION BY RELIABILITY OF COORDINATES DETECTED FROM VIDEOS FOR FINGER CHARACTER RECOGNITION |
|
Author(s):
|
Shinpei Hagimoto, Takuma Nitta, Ari Yanase, Takafumi Nakanishi, Ryotaro Okada and Virach Sornlertlamvanich |
|
ISBN:
|
978-989-8704-26-9 |
|
Editors:
|
Piet Kommers and Pedro Isaías |
|
Year:
|
2021 |
|
Edition:
|
Single |
|
Keywords:
|
Finger Character Recognition, Sign Language, Finger Character, Knowledge Base Creation, Reliability of Coordinates,
Video Metadata |
|
Type:
|
Full |
|
First Page:
|
169 |
|
Last Page:
|
176 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
In this paper, we propose a novel knowledge base creation method by reliability of coordinates detected from videos for
finger character recognition. A sign language communication is a crucial method for deaf or hearing-impaired people.
One of the most important issues is the preparation of video as a training data set for finger character recognition.
Unfortunately, preparing a large enough video data set for finger character recognition is a labor intensive and
time-consuming task. Instead, we are proposing a finger character knowledge base consisting of video data and the
designed metadata. Our method enables the creation of a knowledge base and metadata with high accuracy from a small
amount of training data. By realizing our method, we can obtain a high accuracy metadata for finger character to use in
the knowledge base. As a result, our proposed method can realize an accurate and robust finger character recognition
system by consulting the knowledge base. |
|
|
|
|
|
|
