Bat calls dataset for deep learning classification

This dataset contains ultrasonic bat call recordings used for CNN-based species classification. It includes: - Filtered recordings from the ChiroVox database, processed using an adaptive spectral subtraction method. - Unfiltered recordings from the Xeno-Canto database. The dataset focuses on six bat species: - Rhinolophus hipposideros - Myotis daubentonii - Eptesicus serotinus - Nyctalus noctula - Pipistrellus pipistrellus - Pipistrellus pygmaeus The data were collected and processed in the context of an academic project at INSA Rennes.