CityNet - a neural network for urban sounds
CityNet is a machine-learned system for estimating the level of biotic and anthropogenic sound at each moment in time in an audio file.
The system has been trained and validated on human-labelled audio files captured from green spaces around London.
CityNet comprises a neural network classifier, which operates on audio spectrograms to produce a measure of biotic or anthropogenic activity level.
More details of the method are available from the paper:
Alison J Fairbrass, Michael Firman, Carol Williams, Gabriel J Brostow, Helena Titheridge and Kate E Jones
An overview of predictions of biotic and anthropogenic activity on recordings of London sounds can be seen at our website londonsounds.org.
The system has been tested using:
anacondadistribution (see here for download details).
Run the following commands to get suitable versions of the required libraries:
pip install -r https://raw.githubusercontent.com/Lasagne/Lasagne/v0.1/requirements.txt pip install Lasagne==0.1 pip install --upgrade https://github.com/Lasagne/Lasagne/archive/master.zip pip install nolearn pip install librosa pip install easydict pip install tqdm pip install git+git://github.com/mdfirman/[email protected]
For training and testing we used a 2GB NVIDIA GPU. The computation requirements for classification are pretty low though, so a GPU should not be required.
How to classify a new audio file with CityNet
python demo.pyto classify an example audio file.
- Predictions should be saved in the folder
- Your newly-created file
demo/prediction.pdfshould look identical to the provided file
demo.py should allow you to classify your own audio files.