word_recognition

Inspiration

Kaggle chellenge by speech recogniton short words each in one wav file

What it does

We do predicton of the word speeched a human.

How I built it

I deside similar words in files have similar spectrograms

Challenges I ran into

The worst problem to find a lightweight architecture of neural network and calculate future accuracy.

Accomplishments that I'm proud of

What I learned

What's next for word_recognition