Convolutional neural network and data used for applied soundscape classification with Soundscapes 2 Landscapes (S2L)

This repository documents the ABGQI-CNN manuscript (DOI: https://doi.org/10.1016/j.ecolind.2022.108831). It contains supplementary materials, data used to train a soundscape classification convolutional neural network (CNN), and data to generate manuscript results. The accompanying code can be found at https://doi.org/10.5281/zenodo.6038459. Files include:

• ABGQI-CNN.tar: saved CNN model weights for the 5-class soundscape classifier using a MobileNetV2 architecture pre-trained with bird vocalization data.
• ABGQI_mel_spectrograms.tar: spectrograms used for fine-tuning the pre-trained CNN, above, with training, validation, and testing data splits.
• freesound_licensing.csv: file names and license information related to Freesound auxiliary files.
• RavenLite_Training_Data_Collection.pdf: a manual for RavenLite ROI annotation.
• S2L_site_geog-env_data.csv: environmental and geographic data (sans GPS) related to site locations in S2L project 2017-2020.
• site_avg_ABGQIU_fscore_075_daytime.csv: the average site rate of soundscape components for 5 a.m. to 8 p.m.
• site_by_hour_ABGQIU_fscore_075.csv: the average hourly site rate of soundscape components
• site_classifications_beta075.tar: a directory containing a CSV for every site with threshold optimized classifications for each 2-s Mel spectrogram
• site_prediction_probabilies.tar: a directory containing a CSV for every site with ABGQI-CNN probabilities for each 2-s Mel spectrogram
• Supplementary_Materials.pdf: includes additional material and analyses related to the accompanying manuscript. 

Contact Colin Quinn at cq73@nau.edu for questions related to this repository or if you have an interest in the original wav recordings. Please be aware that underlying software, specifically for the CNN implementation, may not continue stability as python libraries are updated.