I developed the Audio Signal Processing Project for an interview. The company gave this assignment as an assigment. I coded this project in about 38-40 minutes and created the dataset. I created the algorithm in about 7 minutes. The algorithm is very performant and very strong against overfitting. Therefore, the software correctly predicted all of the test data without overfitting. The algorithm that I developed has a total of 8114 neurons (total of input and output layers) and an average of 2 million parameters.
In my Audio Signal Processing Project, I gave importance to data cleaning by removing noise while processing data. In this way, the data focused only on the specified sound. In this software, I distinguish between walking and speaking sounds. The software that I made detected how a person makes a sound while walking.
Kind regards, Emirhan Bulut
You can download this model as using this link. (Emirhan_Model.zip) : https://github.com/emirhanai/Audio-Signal-Processing-Artificial-Intelligence-Project/releases/tag/model
###The coding language used:
Python 3.9.8
###Libraries Used:
Tensorflow
NumPy
Shutil
Keras
os
pathlib
IPython
Name-Surname: Emirhan BULUT
Contact (Email) : emirhan@isap.solutions
LinkedIn : https://www.linkedin.com/in/artificialintelligencebulut/
Kaggle: https://www.kaggle.com/emirhanai
Official Website: https://www.emirhanbulut.com.tr