Freelance Audio Annotator – Transcription and Emotion Recognition
I annotated audio clips for subtitles, distinguishing speakers and recognizing emotions in challenging, noisy Kenyan accents. My responsibilities included transcribing spoken dialogue accurately and identifying emotional nuances for sentiment modeling. This work aided in training AI systems for improved speech recognition and sentiment analysis in local dialects. • Demonstrated native Swahili fluency during annotation. • Used oTranscribe and Google Docs for transcription. • Applied sentiment analysis to enhance label quality. • Operated independently with a focus on data quality.