/Instance-Segmentation-of-Neural-Cells

Instance Segmentation of Neural Cells using Mask RCNN

Primary LanguageJupyter Notebook

-Instance-Segmentation-of-Neural-Cells

Instance Segmentation of Neural Cells using Mask RCNN Objective: The objective of this project is to familiarize ourselves with the process of training the Mask RCNN model for instance segmentation of neural cells. We will utilize the Sartorius - Cell Instance Segmentation dataset, which is part of a Kaggle competition aimed at segmenting neural cells present in microscopic images.

Dataset: The Sartorius - Cell Instance Segmentation dataset can be obtained by registering on Kaggle through the following link:

Project Description:

  1. Mask RCNN Training with CSPResNet50 Backbone:

Note: The Mask RCNN model, augmented with the CSPResNet50 backbone, will be used for instance segmentation of neural cells in the provided microscopic images. The final report will include the model's performance evaluation, along with segmented images as visual examples. This project will serve as a valuable resource for understanding the Mask RCNN approach in instance segmentation tasks, particularly in the context of neural cell analysis.