A new product offered by Skrybot – speech recognition using the HTTP protocol.

We are happy to present our new product called WebSkryBot. It is a server version of a speech decoder working with Linux/Unix and Windows operating systems. The speech recognition server is accessed online through the HTTP protocol and makes it possible to add a speech recognition functionality to one’s own applications without having to run a speech decoder on a client platform. The software can be used for the purposes of mobile, browser-based applications written in any programming language supporting socket-based communication.

Features of the new technology:

• a simple communication protocol based on sessions and on the HTTP protocol,
• decoding results are sent via a return XML with accurate time stamps of the decoded words and the confidence level is provided,
• the server which is decoding speech works in a multi-thread mode, dividing one acoustic and language model among all the threads,
• the decoding sessions are filed and can be seen using an internet browser,
• the communication with the server is socket-based so there are no major limitations concerning the programming language in which the client application is written.
• We create individual language corpora with impact on recognition of chosen phrases and words on demand.

If you want to implement speech recognition in your product, contact us and get the client program source code that demonstrates how to connect to our speech servers. Thanks to this you will be able to find out how to use the technology while having decoding hosted by us or on your servers.