{"paper_id":"0f3cf509-8fb9-40ea-8f9b-d41f1e73de80","body_text":"This is a Preprint and has not been peer reviewed. This is version 1 of this Preprint.\nYou must log in to post a comment.\nThere are no comments or no comments have been made public for this article.\nThis is a Preprint and has not been peer reviewed. This is version 1 of this Preprint.\nAdd a Comment\nYou must log in to post a comment.\nComments\nThere are no comments or no comments have been made public for this article.\nAutomated bird sound classification plays a critical role in biodiversity assessment, ecological monitoring, and conservation research. Many current approaches use single-resolution spectrograms, which fail to fully capture the multi-scale acoustic features of avian vocalizations. We present DualStack, a new method that vertically stacks high-resolution and low-resolution Mel spectrograms into a single image, allowing convolutional neural networks to jointly learn fine temporal and broad spectral patterns. Using a dataset of 967 recordings from 22 species sourced from Xeno-Canto, DualStack achieved 86.63% classification accuracy, outperforming both single-resolution baselines and a BiParallel ResNet18 multi-branch architecture. This method improves species identification accuracy while remaining applicable to real-time monitoring, supporting more effective conservation efforts and large-scale ecological studies.\nhttps://doi.org/10.32942/X2TD2Z\nEngineering\nbird sound classification, bioacoustics, Biodiversity monitoring, ecological informatics, species identification, conservation technology, acoustic monitoring, spectrogram analysis, machine learning\nPublished: 2025-08-11 06:53\nLast Updated: 2025-08-11 06:53\nCC BY Attribution 4.0 International\nLanguage:\nEnglish","source_license":"CC-BY-4.0","license_restricted":false}