UrbanSound - QHacks2019 Project - Github
Model Training: https://github.com/simonkrol/QHacks2019
Create a virtual environment to house the project (optional):
virtualenv -p python3 UrbanSound
cd UrbanSound
Clone the repository:
git clone https://github.com/simonkrol/QHacks2019-app.git
Install dependencies:
pip install -r requirements.txt
Since the model is too large to host on GitHub, it is not included in the repository. If you'd like, you can download the model I trained, from my Google Drive (I've also included a few test files)
Start the server:
python app.py
Either send a request through the requests library, postman or access the webpage endpoint at http://localhost:5000
The result values are expressed as a list of Strings containing the events occuring throughout each part of the sound file.
Sirens from 0 until 3 seconds
Dog Barking from 4 until 9 seconds
No Identifiable Sound from 9 until 12 seconds
Gun Shot at 13 seconds