On October 10th, we upgraded the transcription support application "AmiVoice ScribeAssist" and implemented the latest speech recognition engine "Bi-LSTM Engine".
[Main version upgrades]
- Changed the voice recognition engine to the latest "Bi-LSTM engine". Improved recognition rate of conference audio
Bi-LSTM engine is one of the advanced technologies of deep learning. By predicting and adding future information, processing is performed from a bidirectional time axis. This further improved the recognition rate of conference audio.
[Eligible for free version upgrade]
Existing users can use the latest engine free of charge.
[AmiVoice ScribeAssist overview]
This is a standalone transcription support application that can be used regardless of the type of meeting (online/offline) and the type of web conference system. It supports real-time recognition and batch recognition, allowing you to perform everything from recording conference audio, voice recognition, text editing, and outputting transcribed content in one stop. It is equipped with many functions suitable for use in conferences, such as an AI speaker identification function, a keyword tagging function, and a subtitle display function on Zoom.
*Please see below for product details.
http://www.advanced-media.co.jp/lp/scribeassist/
