Italian , German and French Speech Data Collection
Budget: $10,000 – $20,000 USD
Project Overview:
We are looking for Italian-speaking participants (individuals or teams) to contribute to a high-quality speech data collection project. The task involves recording natural, unscripted conversations between two speakers in a controlled indoor environment using professional audio equipment.
This project is part of a large-scale AI dataset initiative and requires strict adherence to audio quality and recording standards.
Key Requirements:
Language: Italian (fluent/native level)
Participants: Two speakers per recording session
Age Range: 18–45 years
Gender Ratio: Balanced participation preferred
Accent: Neutral or mild regional accent (strong dialects not accepted)
Recording Specifications:
Format: 48kHz, 32-bit, mono
Setup: Dual-channel (each speaker recorded on a separate track)
Equipment:
Condenser microphone (studio quality)
Audio interface (e.g., Focusrite or equivalent)
Environment:
Quiet indoor setting (no echo, minimal background noise)
No external disturbances (traffic, typing, music, etc.)
Task Details:
Record natural conversations (not scripted or read)
Each session duration: 5 to 30 minutes
Multiple sessions required per team
Topics will be provided (e.g., travel, lifestyle, technology, finance, etc.)
Conversations should include natural flow:
Interruptions
Overlapping speech
Real-life interaction
Deliverables:
High-quality dual-channel audio files
Proper file naming and organization
Metadata (participant and session details)
Signed authorization documents
We are looking for Italian-speaking participants (individuals or teams) to contribute to a high-quality speech data collection project. The task involves recording natural, unscripted conversations between two speakers in a controlled indoor environment using professional audio equipment.
This project is part of a large-scale AI dataset initiative and requires strict adherence to audio quality and recording standards.
Key Requirements:
Language: Italian (fluent/native level)
Participants: Two speakers per recording session
Age Range: 18–45 years
Gender Ratio: Balanced participation preferred
Accent: Neutral or mild regional accent (strong dialects not accepted)
Recording Specifications:
Format: 48kHz, 32-bit, mono
Setup: Dual-channel (each speaker recorded on a separate track)
Equipment:
Condenser microphone (studio quality)
Audio interface (e.g., Focusrite or equivalent)
Environment:
Quiet indoor setting (no echo, minimal background noise)
No external disturbances (traffic, typing, music, etc.)
Task Details:
Record natural conversations (not scripted or read)
Each session duration: 5 to 30 minutes
Multiple sessions required per team
Topics will be provided (e.g., travel, lifestyle, technology, finance, etc.)
Conversations should include natural flow:
Interruptions
Overlapping speech
Real-life interaction
Deliverables:
High-quality dual-channel audio files
Proper file naming and organization
Metadata (participant and session details)
Signed authorization documents