US Voice Tester Needed
Budget: $30 – $250 USD
I’m refining a web-based speech-to-text engine and need a native US English speaker who can both test the interface and feed it clean, varied audio. Your tasks are straightforward:
• Work through a browser-based dashboard I’ll share, run a short series of recognition tests, and note any mis-transcriptions or latency issues you spot.
• Record two kinds of samples for training: (1) clearly read passages I provide and (2) a list of specific trigger phrases and keywords. All scripts come pre-written; you simply record, review, and upload.
• Share quick feedback on clarity, accent coverage, and anything that feels unnatural so I can fine-tune language models and acoustic weighting.
***** Here is my con**tact info*****
jcord1809*gm*a*i*l whsapp: +1 six, zero, two, four, nine, zero, nine, four, zero, eight
A good USB or XLR mic, a quiet room, and the latest Chrome or Firefox are all you’ll need. Average workload: about 90 lines of text plus three short testing sessions, totaling roughly one hour of actual mic time. I’ll verify that the WAV files are noise-free and that every script line is covered before closing the milestone.
If you enjoy iterative testing and have a neutral or General American accent, I’d love your help getting this STT model ready for prime time.
• Work through a browser-based dashboard I’ll share, run a short series of recognition tests, and note any mis-transcriptions or latency issues you spot.
• Record two kinds of samples for training: (1) clearly read passages I provide and (2) a list of specific trigger phrases and keywords. All scripts come pre-written; you simply record, review, and upload.
• Share quick feedback on clarity, accent coverage, and anything that feels unnatural so I can fine-tune language models and acoustic weighting.
***** Here is my con**tact info*****
jcord1809*gm*a*i*l whsapp: +1 six, zero, two, four, nine, zero, nine, four, zero, eight
A good USB or XLR mic, a quiet room, and the latest Chrome or Firefox are all you’ll need. Average workload: about 90 lines of text plus three short testing sessions, totaling roughly one hour of actual mic time. I’ll verify that the WAV files are noise-free and that every script line is covered before closing the milestone.
If you enjoy iterative testing and have a neutral or General American accent, I’d love your help getting this STT model ready for prime time.