Unveiling Microorganism Diversity through Neural Network-driven Classification

Le résumé fourni par la source

Our planet has a vast diversity of microorganisms. To better understand these numerous, frequently tiny life forms, this study addresses an automated solution for microorganism classification. The motivation lies in the limitations of traditional classification methods, such as microscopy, which are time-consuming and may lack accuracy due to the abundant types of microorganisms. Therefore, whether deep learning, can provide a more effective answer for microorganism classification is the primary focus of this study. This research uses the EMDS-6 dataset, we applied preprocessing techniques, including data augmentation and image normalization, to enhance dataset quality. Our results are promising, with the proposed CNN-based model Inception v3 achieving a remarkable testing accuracy rate of 93%, signifying the potential of deep learning to significantly enhance microorganism classification. This research bridges modern deep-learning technology with biological sciences, offering valuable applications in areas like environmental monitoring, disease detection, and pharmaceutical research.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.