Search code examples
androidiosspeech-recognitiontext-to-speechspeech

Which speech recognition system to use?

I want to implement following idea. I need my user to open web page or mobile application and say something. The speech of user will be recorded end sent to a local server, and I want the server to process it and generate some voice output.

My questions are next:

  1. Which Speech Recognition API to use? In my case audio file will be sent to server, and after that it will be processed.

  2. I need to recognize names and last names. This is a potential problem, because I don't live in English-speaking country, and names and last names could be rather strange to English-speaker. I want to implement this by adding required names and last names to "dictionary" of speech-recognizing system, so I need one that can expand its dictionary.

  3. I need a free decision.

So, which Speech Recognizing API should I use in implementation of my idea?

Solution

  • I have had success in the past using the CMU Sphinx toolkit for speech recognition. I had used a variation of it within in Android application. If I remember correctly it did allow the creation of a dictionary of specific words to look for so it may fit your needs.