I need help. I have no idea were to start.
I am trying to make a small device with a microphone and a screen get audio input with a microphone and turn it into text on a screen, a couple inches big. The only languages i know how to program are in C and C#. I have never programmed a chip of any kind before.

Thank you for the help.

 

Dude I didn't get it. Please explain in details what you want.

 

There is a box

There is a microphone and a screen on the box.

There is also a circuit board in the box.

When i talk into the microphone, it will go through the circuit board.

the curcuit board will "translate" my words into text using speech-to-text

Then it will display it on the screen

i need help with my box


It is like on cell phones, where you say somthing like "Call Bob" and it calls bob, exept my box is a small screen with text and a micophone.

 

What do you imagine the circuit board does? Are you at all conversant with commercial applications that do this, like Dragon "Naturally Speaking"?

 

I am bad with comercial products and electronics, but that is why i am here, i am trying to learn. I am just a hobbyist.

I was thinking about some kind of hardware or microcontroller on the cuircuit board that had speech to text on it. Maby one microcontroller with microsoft .NET micro and another one of these things : http://www.sensoryinc.com/products/RSC-4x_series.html

I am not asking anyone to do any reaserch or design anything. Just asking people with experience on this topic to give me a word or two about the main concept of what to do. Like what are some good programs to use, maby a couple good chips that work well, any good books, a website with information about this or anything else that would help me get started.

thank you

 

It's a long list. You need to know how mechanical energy gets turned into electrical energy in a microphone.

The microphone signal is very small, so learning about amplification is useful

For a computer to deal with an electrical signal, that signal must be converted into a series of numbers by an analog to digital converter.

The number stream must be recognized by the computer as speech. So analyzing characteristics that can be used to separate speech from noise is important. Not all speech is alike (think of regional accents and vocal pitch), so some means must be found to assign words to the interpreted speech.

You can probably find out more by Googling "speech to text" and going through the 20,100,000 hits - http://en.wikipedia.org/wiki/Speech_recognition