Ornithology, the study of bird species and their vocalizations, is vital to our comprehension of biodiversity, ecosystem health, and avian population conservation. Rich information found in bird songs and calls can be utilized to identify different species, comprehend their behavior, and monitor migratory patterns. However, manual observations and expert knowledge are crucial components of traditional methods of bird identification, which are time-consuming and frequently unfeasible for large-scale investigations. New paths for automatic bird species recognition have been made possible by recent developments in machine learning and audio processing technologies. These technologies can be used to create systems that can recognize different species from audio recordings and evaluate vocalizations made by birds. Capacity not only improves ecological research but also offers useful instruments for public participation and conservation initiatives. Our goal in this study is to create an auditory recognition model for birds that can discriminate between Barswa, Comsan, and Eaywag-One, three distinct bird species. These species were chosen because they have unique vocal traits and a large dataset was available. Convolutional neural networks (Permana et al., J King Saud Univer–Comput Inf Sci, 2021), which have demonstrated remarkable performance in image and audio recognition applications, are the foundation of our methodology. We use an extensive pipeline comprising feature extraction, data cleaning, model training, and assessment.