Request for information - unusual DATA request - Japanese
We are exploring a potential voice data collection project in Japan, focusing on family relationships (parents and siblings) . The data will be used for speech recognition or voice model development.
Target Participants
A total of 100 family pairs (200 individuals) consisting of:
-
Parent Child pairs:
- Father + son (male child, age 14+)
-
Mother + daughter (female child, age 14+)
-
Sibling pairs:
- Brothers (male + male, 14+)
- Sisters (female + female, 14+)
- Mixed siblings (male + female, 14+)
Each category includes 20 pairs , totaling 100 pairs (200 participants) .
At least half of the child participants must be aged 14 or above .
Geographic Coverage
Participants must be from across Japan , distributed as follows:
-
Kanto 30 pairs
-
Kansai 30 pairs
-
Hokkaido/Tohoku 10 pairs
-
Chubu 10 pairs
-
Chugoku/Shikoku 10 pairs
-
Kyushu 10 pairs
️ Total = 100 pairs
Recording Requirements
Each participant must record:
-
1 trigger word pronunciation
-
10 free-form utterances (spontaneous speech)
-
Reading aloud (~2 minutes of text)
Technical specifications:
-
Environment: Quiet indoor space or soundproof room ( Equipment: High-quality lavalier microphone placed near the mouth
-
Format: WAV , 48kHz, 16-bit, uncompressed
️ Project Nature
-
Status: Request for Information (RFI) not yet confirmed
-
Objective: Assess feasibility, potential resources, and anticipated challenges
-
Deliverable: Feedback on availability, capacity, and risks
️ Potential Challenges (to assess)
-
Recruitment difficulty: Finding genuine parent-child or sibling pairs in required age ranges across all Japanese regions.
-
Recording logistics: Ensuring controlled acoustic environments nationwide.
-
Equipment consistency: Participants access to or provision of quality lavalier microphones.
-
Compliance: Managing consent and privacy for minors (1417).
-
Cost and coordination: Travel, scheduling, and regional distribution management.
In short:
This is a Japanese family voice recording RFI involving 100 family pairs (parents and siblings) from various regions of Japan , each recording structured and free speech using high-quality audio standards. The client is seeking to confirm whether vendors can recruit, record, and deliver this dataset and to identify any feasibility concerns before proceeding.
Originally posted on Himalayas