How to create a Wake-word solution for Voice Activation/Voice Modelling

TensorFlow
1

What we require:

Wake-word solution for the phrase “Hey Louie” i.e. whenever user says “Hey Louie”, Louie Voice Control (LVC) app should get activated

Critical requirements of model:

  1. Should work on both Android & iOS
  2. Model should be trainable from scratch (All code to be shared with us)
  3. Model size between 5~10 MB
  4. Fully functional production ready model required

Cost of development/training will be considered.