Based on the work of Michael T. Johnson,Narjes Bozorg, Sidrah Liaqat, Neenu Jose, Patrick Conrey, Anthony Tamasi Original work here: https://github.com/UKYSpeechLab/ukybirddet

Required Software:

• Python version 3.6
• Tensorflow-gpu (tested on v1.8)
• Keras
• other packages and dependencies listed in requirements.txt

Project directory structure

For running this project on the DCASE challenge 2018 data, follow this directory structure: (unzip the provided compressed files in the main project directory)

• < project directory >/adaptation_files
• < project directory >/labels
• < project directory >/prediction
• < project directory >/trained_model
• < project directory >/workingfiles

The directory 'workingfiles' has four subdirectories that contain feature files.

• < project directory >/workingfiles/features_baseline (download link)

In order to reproduce the results of this submission, place the python files in the main project directory and run in the following order:

• birddet_baseline.py