2
sunny44
22d

Hey Guys,
I want to build a voice assistance like Ok Google from scratch using ML.
Actually, I'm unable to figure out How can I achieve this 😅.
I want App like When I give command like Open DevRant then the App should work like Google App.
Ok all Good.
But I want to know If I write a code to open any App like Open APP_NAME
then It is coded by me not my app is learning this.
Sorry If I'm unable to Explain this to you.
I want to know Should I have to code all procedures for doing task like open any app, calling any number, etc myself or is there any way that my App can learn on its on 😔.
If somebody understand this please suggest me what is best for this.

Comments
  • 1
    What device are you targeting? Programming language?
  • 0
    @ScriptCoded I'm creating this for myself so Windows and Androids 😊
  • 1
    @ScriptCoded Ohh and I want to use Tensorflow
  • 0
    Audio -> Overlapping frames -> Phoneme classification -> Phoneme sequence matching

    Traditional approach, low accuracy, works (not always)!
  • 0
    @sudocode What about Commands?
    Should I have to Hard coded them?
    Like Open Any App, Play Music, Etc
  • 0
    Every known english word has a phoneme sequence. Also, Phoneme sequences can be created for unknown words as well ( like devrant).

    Ofcourse this is a very high level perspective of a very traditional approach but hope you got the idea!
  • 0
    Wtf does that even supposed to mean.
Add Comment