Combining Crowdsourcing and Machine Learning for Image Segmentation and to Define Cell Type

To combine crowdsourcing and machine learning to create a new platform for image processing. Also, to develop a new conceptual framework for understanding the notion of cell type and how to manipulate it using sequencing and perturbations.

In this project, the Raj lab developed software for preparing images for crowdsourcing and averaging the results of the crowdsourced data, Quantius. Code and instructions for running machine learning segmentation using Quantius crowdsourced results can be found on their open wiki page.