Software tools to build deep learning microscopy segmentation and analysis models with less training data. Pretrained MicroNet encoders are available for download. Leverages transfer learning from classification models trained on a large (>100,000 images) dataset of microscopy images.
References explaining about the Dataset : https://www.nature.com/articles/sdata2018172 The dataset can be downloaded from here : https://b2share.eudat.eu/records/80df8606fcdb4b2bae1656f0dc6db8ba
Run main.py for training the custom CNN model for the image classification Run pretrained_model.py to load pre-trained Resnet model and finetune for the respective task.