Speech Recognition API

Question

I need to automatically decrypt some short MP3 files as part of the proof of concept that I'm working on. I am currently browsing for cloud solutions or web API services to send MP3s as a simple HTTP request and receive transcription.

The only free open source solution I found here , but the demos don't seem to work (at least not for the files I need to decrypt). I found some corporate solutions for call centers, but so far I can’t just integrate anything into the project.

Are there any speech recognition services on the website? One that can filter out a little noise will be a plus.

Answer 1

Here is an unofficial method to access ASR features for Google. I just tested yesterday and it still works - you can get a JSON style ASR output with words and a related confidence score with FLC audio sampled at 16 kHz.

Answer 2

This may be a good match. In addition, their techcrunch profile ( See This ) lists competitors as: SimulScribe, SpinVox, Vlingo, Nuance, Microsoft, Google Some of these links may be useful.

Vlingo, Bing, and Google have cloud recognizers, but I don’t think they make them public. I believe that they are available only from their authorized customers.

To prove the concept (and small volume), have you just considered using the desktop speech engines that come with Windows 7? What is the difference between System.Speech.Recognition and Microsoft.Speech.Recognition? may be helpful. MS descriptor parsers come with a dictation grammar, and it looks like this is what you need.

Answer 3

You can also try the Windows 7 speech recognition engine for creating subtitles. Here is a tool for this.