Danish bat dataset - PAMGuard deep learning tutorial

Deep learning and bats - PAMGuard tutorial dataset

General Information

Overview

This is a dataset to be used in part two of the PAMGuard tutorial *Deep learning in PAMGuard*. The dataset is from an AudioMoth recording near a woodland in Denmark at 384kHz which will record the sounds (echolocation and social) of all species of bats in the area.

Year of data collection

2020 (AudioMoth), 2017-2020 (deep learning training data)

Geographic location of data collection:

Denmark

Funding sources that supported the collection of the data:

Carlsberg Semper Ardens Research Project and The Independent Research Fund Denmark

Recommended citation for this dataset:

Brinkløv S. M. M., Macaulay, J., Bergler, C., Tougaard, J., Beedholm, K., Elmeros, M., Madsen, P. T. M. (2023) Open-source workflow approaches to passive acoustic monitoring of bats. Methods in Ecology and Evolution Signe: https://doi.org/10.1111/2041-210X.14131

Data & File Overview

Description of dataset

These data include a small sample of acoustic recordings of bats at woodland edges made using an AudioMoth running firmware version 1.4.4 (Hill et al., 2018, openacousticdevices.org). 

The full dataset that these data are taken from can be found [here](https://doi.org/10.5061/dryad.4xgxd25fh) and the associated publication is [here](https://doi.org/10.1111/2041-210X.14131);

Files

The files are single channel .wav files made on an AudioMoth. These are 16 bit files but the recorder ADC was 12-bit and therefore the dynamic range is around 60 dB. The files are located in the wav folder. The deep learning model is located in the root directory. It is a .pk file which can be imported directly into PAMGuard. 

The folder structure is as follows:

```md├── wav                                            #Folder of .wav files│   ├── 20200817_010002.WAV│   ├── 20200817_010012.WAV │   └── '''├── README.md                                        #This readme file├── BAT_MULTI_JAMIE_5ms_256fft_10hop_MM_10_90_128_256_JIT_PITCHTIMECHANGE_AUG_V1.pk     #Bat deep learning model in PAMGuard compatible format.└── ```

Device locations

Odense (‘Skovsoen’): N55.3741; E10.3819

Sharing and access

These data are open source under Creative Commons Attribution 4.0 International. This means that these data can be used in other tutorials as long as the original authors are credited. Using these data for scientific purposes requires the permission of the authors.