Using machine learning to build a complementary sight for visually impaired.
- python 3.x
- pickle version 4.0
- Google Colaboratory (GPU)
- clone this repository and upload to google drive
- open the Jupyter Nootbook file with Google Colaboratory and
Edit->Nootbook Settings->select GPU under Hardware accelerator Testing_Datasetconsists of original images that we took, background removed images and cropped images processed withTesting image segmentation.ipynbprimary_model.pkldownload our trained model from drive and make sure this file is saved in your working directorytesting.ipynbrun all in the file
AI-eye description.pdf