Welcome to DeepLearnMOR
DeepLearnMOR (Deep Learning of the Morphology of Organelles) is a deep learning framework that can rapidly classify a diverse array of organelle morphological abnormalities based on fluorescence images.
Currently, DeepLearnMOR classifies morphology of three major energy organelles in Arabidopsis, including chloroplasts, mitochondria and peroxisomes. You can download the fluorescence image dataset to play with it.
Deep Learning Models
DeepLearnMOR consists of Transfer Learning models and Convolutional Neural Network (CNN). The feature visualization components in DeepLearnMOR identify and extract key features used for decision-making in classification thus provide model interpretability. You can check the models, please feel free to “Deep Learn More” by trying it out.
DeepLearnMOR would be expanded to classify organelle morphology beyond plants to human cells and other systems. This framework can potentially be deployed in research and clinical labs, microscopic facilities and cloud-based distributed clusters for large-scale and real-time applications.