Web demos Furnariidae recognition


Android App | Source code: (1) | Data: {1} | Related publications: [1]

This web-demo allow the automatic classification of Furnariidae species from Litoral Paranaense region using speech-related features and machine learning. The  system can be tested using the example register or uploading some register. The preprocessing can be set to use or not: Wiener-based filter and acoustic activity detector. Then, after the feature extraction process, the sample is classified using the best model trained using all the reported data.
Results (available to download):
– Processed audio file,
– The graphical results of the audio files and their spectrograms,
– The extracted features file (arff format),
– The trained model (WEKA format),
– The label of the recognised species.