The problem Detector Surge solves

Traditional method of analyzing MRI by Doctors is used which is time consuming and cause Human error, so to overcome these problems an alternative way is to design a system that will automatically identify the presence of tumor in MRI images using deep learning and neural networks. This system would also be able to provide faster and accurate solutions as compared to the conventional system.

Challenges we ran into

Faced Diffilculty in finding a appropriate Data-set with large number of images
As a newbie to Machine learning domain it is very difficult to work with huge number of funtions and libraries
As we have used MobileNet architure so most of the common inbuilt functions are not working , we have encounterd with number of errors , So finding new functions and there respective libraries is quite a troublesome . For ex ( A function names model.predict_classes() work fine for simple model but not for MobileNet architucture
so we founded a new function named model.predict() and then used a different logic so it can behave as model.predict_classes() )
Faced Pixel Related issue for showing false result for Feed images