Generative Audio Evaluation - French (France)
We are looking for Generative Audio Evaluation Specialists! A great entry point into ongoing work within one of our most active AI markets!
Job Type: Freelance
Location: Remote (France)
Work Schedule: Part-time - 10+ hours per week. Flexible - work whenever you want!
Rate: 16 USD/hour
Help Shape the Future of AI
We’re currently hiring Generative Audio Evaluators to join a highly specialized evaluation project focused on testing and assessing the performance of a cutting-edge generative audio model. Your work will play an important role in identifying performance issues and providing feedback that can help improve the model.
What You’ll Be Doing
You will evaluate AI-generated audio for short video and text-to-audio tasks, following detailed project guidelines. Your main responsibilities will include:
Audio Quality & Naturalness: Assess speech, background music, and sound effects for clarity, quality, and naturalness.
Audio-Video Alignment: Evaluate synchronization between the generated audio and video, including lip-sync and timing.
Caption Compliance: Check whether the generated audio accurately follows the provided text captions and task requirements.
Voice Quality: Evaluate whether generated voices sound natural and, where applicable, match the reference audio.
Evaluation & Feedback: Assign accurate scores and identify issues according to the provided evaluation guidelines.
What We’re Looking For
Native-level fluency in French
Strong attention to detail and excellent listening skills
Ability to follow detailed evaluation guidelines consistently
Good understanding of spoken language, audio quality, and context
Reliable equipment — high-quality headphones are required
Ability to complete the evaluation test within 48 hours of receiving the assignment
Qualification Process
- Application (You are here!)
- Screening Survey
- Language Knowledge Assessments
- Target Language
- English
- RWS Vendor Onboarding
- Internal Testing
- Project-specific testing will be required, such as TTS Evaluation, Audio Evaluation, or Transcription Testing.
Cette annonce provient de Lever. Voir l'annonce originale ↗