I tried using docomo speech recognition API and Google Speech API in Java
When I checked the library etc. in order to install the function of making the product at school perform voice recognition and converting it into a character string, I used it because there was an API for docomo and Google.
Preparation
docomo Speech Recognition API To use Google's voice recognition API, you must join the following groups Chrome-dev Google Cloud Console
Please create (sign in) an account on docomo and the Google Cloud Console site. Also, in Google Cloud Console, create a new project and enter ** "speech" ** from the API library search window to enable the API. At docomo, when you apply for an API Key, it will be issued in a few minutes, depending on the time of day.
(Already enabled in the image, but for reference. Joining the Chrome-dev group brought up this API)
Execution result
The upper string is docomo and the lower is Google (I'm sorry for the strange display because it was a program in the middle of the experiment)
The source code has been uploaded to GitHub for your reference.
It's a Gabagaba program, but please look at it with warm eyes as you are still studying. Google voice recognition is taking out the first one. There are about 3 candidates for the response body, so please try to modify them yourself. We also added a program to play the recorded voice, so if the recognition does not work, check the voice and check if there is any noise.
If you have any typographical errors, please leave a comment or edit request.
Recommended Posts