Experiences in data collection for the training of an automatic speech recognizer in Sepedi

被引:0
作者
Manamela, MJD [1 ]
Botha, EC [1 ]
机构
[1] Univ North, Dept Comp Sci, Sovenga, South Africa
来源
2002 IEEE AFRICON, VOLS 1 AND 2: ELECTROTECHNOLOGICAL SERVICES FOR AFRICA | 2002年
关键词
D O I
10.1109/AFRCON.2002.1146866
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
This paper reports on the process, experiences and statistics of collecting Northern Sotho (Sepedi) telephone speech data for the training of an automatic speech recogniser. A toll-free telephone speech recording platform was used for the data collection. An analysis of the collected speech data together with observations made during supervised calls are included before discussing preliminary results.
引用
收藏
页码:377 / 381
页数:5
相关论文
共 11 条
[1]  
DENOS EA, 1994, P 18 U AMST, P87
[2]  
LOUW PH, 2001, P 12 ANN S PRASA FRA, P115
[3]  
LOUW PH, 2001, 08 U STELL AFR SPEEC
[4]  
LOUWRENS LJ, 1995, N SOTHO
[5]  
MANAMELA MJD, 2001, P PRASA FRANSCH S AF, P143
[6]  
MANAMELA MJD, 2001, P S AFR TEL NETW APP, P684
[7]  
MATSUI T, 1999, EUR 99 BUD HUNG 5 9, P2283
[8]  
POULOS G, 1994, LINGUISTIC ANAL N SO
[9]  
ROUX JC, 2000, 2 INT LANG RES EV C, P975
[10]  
SANDERS E, 2001, P 12 ANN S PATT REC, P104