Bioacoustic classification in the age of deep learning: A survey of methods and applications

Bioacoustics is the scientific study of sound production and reception in animals. These vocalizations are fascinating objects of study for biologists and ecologists because they can provide insights into animal behavior, as well as information about issues such as global warming, climate change, and agriculture. However, manually categorizing bioacoustics properties can be very hard. This is where automated bioacoustics sound classification comes in, which has become increasingly important due to advances in computing and the availability of affordable digital sound recording devices. In this paper we will examine various deep learning methods that have been used to classify animal sounds, including those of birds, wolves, and marine animals.