November 1, 2013 at 15:21Eleven - service for automatic translation of speech into text
Hey. We, Borya and Polina, will talk about our Eleven project.
We are developing Eleven - a service for automatic translation of speech into text and a plugin for Skype. The service contacts Skype during a call, processes a voice conversation and returns it to the user in text form within 5-7 seconds. The text log is divided into familiar blocks: your words and the words of your interlocutor. Communication with Skype is needed to determine the beginning and end of a conversation.
The idea of speech recognition is not new, and the world is actively developing voice dialing services with very diverse functionality. Our product is intended for freelancers - developers and designers - and is aimed at improving the convenience in their work.
Everything is very simple: we cut your voice and send it to the server via HTTP. We get the answer back and save it in the cloud. The plugin itself connects to Skype and monitors the beginning and end of the conversation. But we think it’s better to see once.
Unpleasant situation:
- Work requires changes.
- And before it was impossible to say?
- I told.
- Proof?
Freelancers often face the problem of a forgotten task. After showing the result to the customer, it turns out that some details are not accurate. The fact is that an unprepared person can keep in memory no more than 3-4 points. Little things in time are not kept in memory at all. Eleven solves a very urgent task - captures the fact of negotiations.
Oh yes. The video we watched through the holes, which we show at all presentations, gave us the name.
www.youtube.com/embed/G2Y0oqZOyl0
We are developing Eleven - a service for automatic translation of speech into text and a plugin for Skype. The service contacts Skype during a call, processes a voice conversation and returns it to the user in text form within 5-7 seconds. The text log is divided into familiar blocks: your words and the words of your interlocutor. Communication with Skype is needed to determine the beginning and end of a conversation.
The idea of speech recognition is not new, and the world is actively developing voice dialing services with very diverse functionality. Our product is intended for freelancers - developers and designers - and is aimed at improving the convenience in their work.
How it works
Everything is very simple: we cut your voice and send it to the server via HTTP. We get the answer back and save it in the cloud. The plugin itself connects to Skype and monitors the beginning and end of the conversation. But we think it’s better to see once.
Why do we do this
Unpleasant situation:
- Work requires changes.
- And before it was impossible to say?
- I told.
- Proof?
Freelancers often face the problem of a forgotten task. After showing the result to the customer, it turns out that some details are not accurate. The fact is that an unprepared person can keep in memory no more than 3-4 points. Little things in time are not kept in memory at all. Eleven solves a very urgent task - captures the fact of negotiations.
Oh yes. The video we watched through the holes, which we show at all presentations, gave us the name.
www.youtube.com/embed/G2Y0oqZOyl0