Bird Species Identification Using Speech Recognition*

This article is significant for its potential to automate and streamline bird monitoring, enabling efficient and widespread species identification for conservation purposes. This study employed deep learning techniques to predict different species of birds based on their vocalizations. To categorize bird species, its been built using a deep neural network model with characteristics collected from bird noises. The dataset utilized for the purpose of training and testing the model comprises of recordings of avian vocalizations obtained from five distinct species, namely “American Robin,” “Northern Mockingbird,” “Northern Cardinal,” “Song Sparrow,” and “House Crow.” Pre-processing of the dataset involved the extraction of MFCCs from each recording, which served as the basis for the features used in training the model.