/FlowersImageClassifierTerminalApp

Primary LanguageJupyter Notebook

FlowersImageClassifierTerminalApp

This repository contains:

  • A jupyter notebook p1_image_classifier.ipynb containing the code to train a model on the oxford_flowers102 dataset
  • And a python script in the folder p2_image_classifier which can be run in the terminal with input arguments:
  • Location of the image of a flower which needs to be predicted
  • Location of model being used (in this project's case you can use my trained model from p1_image_classifier.ipynb which I have named model1.h5 and added in the same folder)
  • Optional Argument 1:The top k classes which were predicted (default value = 1)
  • Optional Argument 2:Location of the JSON file containg class names(flower names) against their indices an example of the input to be used in the terminal: python predict.py ./test_images/orange_dahlia.jpg ./model1.h5 --top_k 5 --category_names label_map.json

Caution:

The model is trained with tensorflow 2.5.0 and other details specified in p1_image_classifer.ipynb, so the SAME requirements are needed for PART 2 of the project

The Terminal Application i.e. p2_image_classifer was run on Jupyter Notebook's Terminal

Don't Forget to change your directory to p2_image_classifier before running the python script