NTT-Tohoku University Speech Data Set for Word Intelligibility Test based on Word Familiarity (FW03)

Data DOI

https://doi.org/10.32130/src.FW03

Producer, Project

Shigeaki Amano and Tadahisa Kondo, NTT Communication Science Laboratory

Yoichi Suzuki and Shuichi Sakamoto, Research Institute of Electrical Communication, Tohoku University

Contents

All words (4000 items of 4-mora-words) contained in the "Word list for word intelligibility test for the hard-of-hearing" and monosyllables.

Speaker

4 trained Japanese speakers (2 males and 2 females); each item uttered once.

Speech file format

WAV format (48 kHz, 16 bit, Mono)

Distribution media

2 CD-ROMs

Licensing

For research and development purposes only

Price

18 000 yen (plus consumption tax for a domestic order)

Note

This corpus was distributed by NTT Advanced Technology Corporation (NTT-AT) until Mar., 2006.

Speech sample for test listening

Word Familiarity 1.0 – 2.5

a.i.kya.ku (アイキャク)

Word Familiarity 2.5 – 4.0

a.to.zu.ke (アドヅケ)

Word Familiarity 4.0 – 5.5

a.i.a.i (アイアイ)

Word Familiarity 5.5 – 7.0

a.ma.gu.mo (アマグモ)