Psst; inflo transcribes with Whisper

Techtel, the creators of inflo, today announced that a new speech to text transcription engine from openai called Whisper, is available to inflo customers who are looking to unlock more potential from their content library.

Whisper has shown that it excels in accuracy and flexibility with results that are exceptional across multiple dialects.  The Automatic Speech Recognition (ASR) system has used machine learning to train and refine it's accuracy from a growing sample of over 680,000 hours of web content.  The process of refining the ASR has made the solution more accurate and robust so that it understands accents, background noise and also technical language.

inflo intelligently indexes content that has been transcribed and aligns time code references enabling keyword searching to navigate the user straight to the relevant clip and exact position in the video where their keyword was spoken.  Additionally, inflo empowers users to browse content with muted audio whilst displaying a script of all dialogue allowing the user to silently read along with their video stream.

For inflo customers who require transcription exports, inflo can generate a Web Video Text Tracks (WebVTT) file separately to export to any 3rd party API connected system.  This tool means that inflo is the answer to update other less agile content management platforms with the transcription files their users are crying out for.

"Transcription is a huge step forward on inflo's artifical intelligence journey" said Peter Stevens, Software Product Manager for inflo.  "we have lots of customers expressing interest in a solution that can provide them with unmetered transcription and we think our approach will be very popular".