ASR and Data Caption Labelling Annotator (Surabaya, Indonesia)

ASR:

Strong Auditory Comprehension: The project's main task was to transcribe audio, some of which had imperfect quality and volume fluctuations. Strong auditory comprehension would help mitigate subjective cases.

Strong Market Language Proficiency: As the source audio was mostly spoken informally, strong knowledge of the market language was essential. This would help with understanding the context of the topics spoken in the audio, as some audio may contain irregularities in pronunciation, such as colloquial language, slang, dialects, and non-standard pronunciations.

Comfort with Sensitive Content: As also stated in the guidelines, some audio might contain extreme profanity or graphic sexual language. Candidates who are unfazed by such content would be a plus.

Caption:

Voice & Character Profiling: Able to assess and categorize speakers based on vocal characteristics, including apparent age group and other relevant attributes. Candidates with the knowledge or experience to differentiate between, for example, a middle-aged man and a teenager based on their voices, with little assistance from visual cues, would be highly beneficial.

Strong Auditory Discrimination: Able to distinguish between multiple speakers, including individuals with similar-sounding voices, based on vocal characteristics and speaking patterns. For example, if a video has three male speakers with somewhat similar voices, it would be helpful if the candidate could determine who is speaking based solely on their voice, especially when there is no clear visual indication of who is speaking