Paper Key : IRJ************810
Author: Vipul Badole,Prathamesh Gawale,Pavan Gawande,Varun Inamdar,A. M. Todkar
Date Published: 15 Nov 2024
Abstract
Bird species identification is essential for biodiversity monitoring, conservation efforts, and ecological studies. Traditional bird identification methods rely on either visual or auditory cues, limiting their accuracy in complex environments. "FeatherScan" presents a survey of advanced techniques combining Convolutional Neural Network (CNN)based models for both audio and image data to improve identification accuracy. By leveraging CNN architectures optimized for audio spectrograms and bird imagery, this study explores the integration of dual-modality models to capture distinct species characteristics. We review current advancements in audio and image recognition for bird species, analyze their strengths and limitations, and propose an integrated approach. FeatherScan aims to provide insights into the potential of hybrid models, enhancing the precision of automated bird identification systems and contributing to the development of comprehensive ecological monitoring tools.
DOI Requested