Cantonese Audio Evaluator
Job Description
The tonal intricacies and unique phonology of Cantonese make it a challenging yet crucial language for advanced AI. As a Cantonese Audio Evaluator, you will be the discerning ear that ensures AI-generated Cantonese speech achieves impeccable pronunciation accuracy, natural intonation, and cultural authenticity, directly impacting AI's ability to communicate effectively with a significant global population.
Key Responsibilities
Systematically evaluate AI-generated Cantonese audio samples for pronunciation accuracy, tonal correctness, and naturalness.
Identify and categorize specific phonetic errors, incorrect tones, and unnatural prosody in Cantonese speech.
Assess the emotional congruence between the AI's vocal delivery and the intended sentiment of the text.
Provide detailed, actionable feedback on speech quality, rhythm, and overall fluency.
Utilize proprietary evaluation tools to mark and segment problematic audio regions.
Participate in calibration sessions to maintain consistent evaluation standards across the team.
Ideal Qualifications
Native fluency in Cantonese, with a deep understanding of its six (or nine) tones and unique phonology.
Exceptional auditory discrimination skills and a keen ear for subtle speech nuances.
Background in linguistics, phonetics, or speech technology, particularly with tonal languages.
Proficiency with audio editing software (e.g., Audacity, Adobe Audition) for detailed analysis is a plus.
Experience with speech evaluation rubrics or quality assurance in voice technologies.
Ability to articulate complex auditory observations into clear, structured feedback.
Project Timeline
Start Date: Immediately
Duration: Ongoing, project-based
• Commitment: Flexible, 15-25 hours/week
Lend your expert ear to shape the next generation of AI voices in Cantonese!