About the Project

This is a project forked from the antiplasti/Plastic-Detection-Model github repository.

A generic image detection program that uses Google's Machine Learning library, Tensorflow and a pre-trained Deep Learning Convolutional Neural Network model called Inception.

Inception was used in the Large Visual Recognition CHallege using data from 2012 and can differentiate between 1000 classes The program applies Transfer Learning to this existing model and re-trains it to classify a new set of images. Essentially, this model can be used to classify any kind of image.

The output is in the form of a probability of what the algorithm thinks the object would be. I have used tkinter for to user to input their image from files and then displayed both the probability and the image as the result.

How to run the code

Make sure you have Python 3 installed, then install Tensorflow on your system, and clone this repo.

Unzip the training_dataset.zip file in the main folder itself. It should look something like this:

with the individual images in each folder

Then, run this cmd in git to initiate the transfer learning:

$ bash train.sh

This installs the Inception model and returns an accuracy from 85-100%.

The training summaries, retrained graphs and retrained labels will be saved in a folder named tf_files.

Finally to run the code, run the cmd:

py app.py

Once the input file is selected, the classifier will output the predictions for each data set. A prediction score between 0.8 to 1 is considered to be optimal.