Acoustic Drone Detection Application
The application processes the signal received from the microphone array. Processing consists of extracting an MFCC or mel-spectrogram depending on the configuration. The extracted feature is fed to a neural network classifier.
Required Python version <= 3.9.19 (Anaconda is highly recommended).
First of all install tensorflow dependencies. Please check the tensorflow installation guide. For optimal performance it is recommended to run the code using a CUDA enabled GPU.
To install requirements:
pip install -r requirements.txt
The code was tested with Python 3.9.19 the following software versions:
| Software | version |
|---|---|
| cuDNN | 8.1.0.77 |
| tensorflow | 2.10.1 |
| CUDA | 12.6 |
To start application run:
python mainwindow.py
This project is licensed under the GNU General Public License v3.0 - see the LICENSE.md file for details.