Generative Audio Evaluation - Dutch (The Netherlands)

Generative audio evaluation

Evaluate AI generated audio in short video and text to audio tasks. Your ratings identify performance problems and provide feedback that can improve the audio model.

Evaluation areas

• Judge speech, background music, and sound effects for clarity, naturalness, and overall quality.
• Check synchronization between audio and video, including lip movements and timing.
• Assess whether generated sound follows the supplied captions and task requirements.
• Review generated voices for natural delivery and, where relevant, similarity to reference audio.
• Assign scores and document issues consistently with the evaluation guidelines.

Requirements

Native Dutch proficiency, strong attention to detail, and excellent listening skills are required. You should understand spoken language, audio quality, and context well enough to make careful judgments.

High quality headphones are required. Candidates must be able to complete the evaluation test within 48 hours of receiving it.

Project terms

This is a freelance project performed remotely from the Netherlands. The rate is $19/hour, with a flexible part time schedule of at least 10 hours per week.