Self-supervised machine learning model for analysis of nanowire morphologies from transmission electron microscopy images

In the field of soft materials, microscopy is the first and often only accessible method for structural characterization. There is a growing interest in the development of machine learning methods that can automate the analysis and interpretation of microscopy images. Typically training of machine learning models require large numbers of images with associated structural labels, however, manual labeling of images requires domain knowledge and is prone to human error and subjectivity. To overcome these limitations, we present a self-supervised transfer learning approach that uses a small number of labeled microscopy images for training and performs as effectively as methods trained on significantly larger data sets. Specifically, we train an image encoder with unlabeled images and use that encoder for transfer learning of different downstream image tasks (classification and segmentation) with a minimal number of labeled images for training.