Speech Recognition Data Annotation (Arabic-English)
A transcription of a 15-minute Arabic audio made for a documentation about the memories of some friends at their old school. The audio contained a single-speaker and was among 1000-2000 words. After Transcribing the audio dataset and doing post-processing and quality assurance procedures, the transcribed text was translated to natural and contextually accurate English and used for video-captioning, taking in consideration using colloquial expressions to match English clarity.