This project has started to classify the 10 most common species of birds from their call. The aim is to develope an application which is able to recognise the most common birds songs. The files are retrived from the https://www.xeno-canto.org/, an opensource website where people can load their recordings.

1. new_retriver.py is used to get the most common birds, thanks to the top_ten class and downloads the files thanks to the class retriver.
2. preprocessing.py is used to process the data and clean them as described in Birds-Song-Recognition-report
3. model.py prepares the data and evaluates the model
4. main_birds.py executes the program.
5. Header.py contains the paths, download it and change the paths inside it with your own ones. This file won't be updated.
6. Birds_Map.html is an html interactive map

Link to Download our DataSet:

• CSV File: https://drive.google.com/file/d/1rrKAuEhpj9qNmZ6IcgIG2_r0HanKim2Q/view
• Parquet File: https://drive.google.com/file/d/183haB0l8z-wro1DYom9ZpyRsfDPJ-W3a/view