Speech Technology Magazine

 

Verbit Launches AI-powered, Real-Time Transcription Solution

AI-driven transcription company introduces a solution for generating speech-to-text through Communication Access Real-Time Translation (CART) technology.
Posted Mar 5, 2019
Page1 of 1
Bookmark and Share

Verbit, a transcription and captioning solution, announced the release of its real-time transcription solution, also known as Communication Access Real-Time Translation (CART) services, which generates live closed captions through its proprietary AI-powered speech recognition technology.

Verbit’s in-house, automated speech recognition (ASR) engine uses adaptive algorithms and a dual process of human transcribers to produce the highest accuracy. The self-learning technology incorporates information from external sources and previous sessions, which are then fed back into the ASR model to continuously train the engine and improve accuracy. The transcripts are also enhanced through a dual layer of professional transcribers, with up to three backup type correctors per session, working in real-time to correct inaccuracies, guarantee precision and ensure reliability.

The CART solution, which is trained to recognize not only commonly used language but also specific terminology and speakers, is currently being used to help bridge the accessibility gap for individuals with hearing impairments and comply with FCC regulations and ADA requirements.

Through its on-demand marketplace of thousands of transcribers, Verbit’s online platform supports easy scheduling, with the ability to book or cancel a session within minutes. The live product also offers unlimited capacity, full support, and provides users the option to receive a downloadable transcript immediately after the session. Additionally, Verbit uses a prorated system that rounds to the minute, not to the hour – going against the standard for transcription services – and doesn’t involve a setup fee.

Page1 of 1