SCRIPTA MANENT.
What is scripta?
Our speech-to-text solution
At Litus AI, we believe that data is one of the main and indispensable assets for any AI company that wants to offer truly innovative solutions.
For this reason, we have invested heavily in creating audio resources, manually annotated and with a strong focus on the Italian language, in various domains of speech: News, Sports, Finance, Technical, ...
Resulting from careful use of these data, Scripta is our speech-to-text service. It is based on a proprietary model capable of providing accurate transcriptions and precise information on who said what and when, with a Word Error Rate (WER) among the best in the market.
What are you waiting for, enrich your systems with an unprecedented transcription service!
What can scripta do?
- Identify the audio regions containing speech
- Recognize and separate the different speakers involved
- Transcribe what was said, with precise punctuation and timing
- Report cases of special communication (laughter, interjections, ...)
What modes does scripta support?
- offline: upload audio files via asynchronous HTTP calls
- streaming: process audio streams via WebSocket
TEST ON FILE
Speaker identification
No file selected