DeepProfiler is a set of tools that allow you to use deep learning and representation learning for extracting phenotypic information from microscopy images of cells.
Links to repositories:
Configuration and downstream analysis notebooks: https://github.com/broadinstitute/DeepProfilerExperiments
Description: Using Reinforcement Learning to streamline/automate the data science process, in particular training image classification models. We believe that creating a well-defined RL protocol will help to speed-up and standardize the process of training machine learning models for scientists.
Links to repositories: https://github.com/broadinstitute/AutoTrain
We are utilizing supervised learning to streamline the process of segmenting nuclei in microscopy images. Creating an efficient segmentation network will save countless hours in the lab because researchers will not have to spend their time manually searching for nuclei when conducting experiments.