When voice contributors provide speech translations of a script, we want to also provide the written transcription of the speech.
Unfortunately most of the ASR (Automatic Speech Recognition) models for African languages cover only a handful of languages and even so most of them aren’t that robust. This limitation of ASR for African languages is one of the motivations for OpenVoice!
So we can’t rely fully on AI for the transcriptions, at least not yet. So that’s were human transcribers come in. You listen to a speech snippet and provide the transcription in the same language.
You should strive to make your transcriptions as accurate as possible and include accents where neccessary.

​
Rewards

Again, here are some ways we plan to Gamify the transcriptions;

Transcription Points: Earn points for each time your transcript is upvoted (meaning its accurate). Extra points if your transcription makes it to the end (selected for corpus). Accuracy Challenges: Participate in accuracy challenges to transcribe difficult voice snippets for additional rewards.

​
Join the community

Join our transcriber community

click here.

VoiceWriters