Inspiration

We couldn't hear others while interacting with other's in the event! So thought we should do something about it.

What it does

It helps to Hear people in a profound way!

How we built it

The code applies a short-time Fourier transform to decompose the audio into its various frequency components. It then detects the frequency patterns that correspond to individual voices. By recognizing these patterns, the code isolates and distinguishes each voice in the audio, allowing the machine learning model being used to separate voices and maybe identify which voice to focus on

Challenges we ran into

Research and While Development It was difficult to find the patterns in voices, so we switched to using machine learning to identify those patterns.

Built With

  • denoiser
  • matplotlib
  • numpy
  • opencv
  • python
  • struct
  • urban-sound-dataset-(https://urbansounddataset.weebly.com/urbansound8k.html)
  • wave
Share this project:

Updates