Skip to content

Latest commit

 

History

History
17 lines (14 loc) · 1.14 KB

File metadata and controls

17 lines (14 loc) · 1.14 KB

Common-disease-detection-using-neural-networks

Language Language

What is this project?

This is an application that is used to detect very common diseases that display their symptoms externally like ey, skin, mouth, nail diseases

How was it achieved?

  • The data was collected from Gogle images using a simple chrome extension called fatkun
  • The data was pre processed and stored as pickles using OpenCV
  • The data was trained using 27 different combinations of Neural Networks and the best model was chose as the one with the least validation using Tensorboard
  • The best model was chosen and reconstructed for the best results.

Tensorboard visualisation

Sample predictions