Generative Audio Evaluation - Chinese (Shanghai)

Role overview Join a specialized evaluation project that benchmarks a cutting-edge generative audio model. This freelance, remote position suits a native Dutch speaker who wants flexible part-time work contributing to AI quality improvements. You'll assess short-form audio outputs and supply structured feedback that helps identify model weaknesses.

Responsibilities - Evaluate AI-generated speech, background music, and sound effects for clarity, naturalness, and overall quality. - Check synchronization between generated audio and accompanying video, including lip-sync and timing accuracy. - Verify that generated audio faithfully follows supplied text captions and stated task requirements. - Judge voice naturalness and, where applicable, consistency with reference audio samples. - Apply scoring rubrics and document specific issues in line with the project's evaluation guidelines.

Requirements - Native-level fluency in Dutch. - Strong attention to detail and sharp listening skills. - Ability to follow detailed evaluation criteria consistently across assignments. - Good grasp of audio quality cues, spoken-language nuances, and contextual interpretation. - High-quality, reliable headphones and suitable home equipment. - Capacity to complete the evaluation test within 48 hours of receiving the assignment.

Benefits and work setup - Freelance engagement, fully remote within China. - Part-time schedule of roughly 10 or more hours per week, with flexible timing. - Hourly rate of 6 USD, paid for time worked on assigned tasks.