Can Voice Signals Be Converted into Matrix Form for Neural Network Training?

AI Thread Summary
Voice signals can indeed be converted into matrix form for neural network training, which is essential for applications like voice recognition. The project is focused on creating a speaker-dependent system, primarily requiring training on the user's own voice. Clarification is sought regarding the specific definition of "matrix form" in this context. Additionally, there is interest in the type of neural network being developed and whether it will be simulated on a PC or constructed as a hybrid analog-digital circuit. This project presents an exciting opportunity for exploration in voice recognition technology.
Madness_hero
Messages
6
Reaction score
0
is it possible to change voice signals to matrix form?
I was thinking of using the neural network to train for voice recognition.
The system I will be building is speaker dependent so I don't think that much training is needed besides my own voice
 
Physics news on Phys.org
I think you'd have to tell us more about what you mean by "matrix form."

- Warren
 
...and what type of neural network are you planning on making? Are you going to simulate it on a PC, or build an analog+digital one in circuit form?

Sounds like a fun project!
 
Back
Top