Hybrid Caption Editing System by Integrating Line Unit Type and Word Processor Type

Kohtaroh MIYAMOTO  Kenichi ARAKAWA  Takashi SAITO 

Publication
D - Abstracts of IEICE TRANSACTIONS on Information and Systems (Japanese Edition)  Vol.J90-D  No.3  pp.673-682
Publication Date: 2007/03/01
Online ISSN: 1881-0225
Print ISSN: 1880-4535
Type of Manuscript: Special Section PAPER (Special Section on Welfare Information and Technology)
Category: 
Keyword: 
accessibility,  captioning,  voice recognition,  digital contents,  voice recognition rate, 

Full Text(in Japanese): PDF(649KB)

Summary: 
There are many digital archive contents available without the benefit of captioning. It is extremely important to caption these contents from audio from the accessibility perspective. Unfortunately, captioning is very much labor intensive. In terms of caption editing methods from voice recognition result, there was "word processor type" but to increase the efficiency for the overall caption editing process, we have introduced "line unit type". "Line unit type" shows high functionality when voice recognition rate is high, but recently it has been found that there are some usability problems when recognition rate is low. In this paper, we have decided to focus on the voice recognition rate. We conducted a preliminary test concerning the voice recognition rate and found that voice recognition rate varies in a high degree. Therefore to improve the efficiency of caption editing we will introduce a hybrid system which integrates "line unit type" with "word processor type". As a result of experimentation for caption editing efficiency against previous caption editing methods, the hybrid system shows advantage in number of interaction and editing time.