Readme.md

DataAnalyzer

To run the project :

• Create a project in pycharm and choose venv environment.

  • Add needed librairies :

    • Application :
      • kivy
      • uvicorn
      • aiofiles
      • opencv-python
      • python-multipart
    • Models :
      • tenserflow
      • keras
      • matplotlib
      • numpy
      • seaborn
      • sklearn
    • Scrapping :
      • Selenium
    • Sounds :
      • gtts
      • playsound 1.2.2 using pip install playsound=1.2.2
      • google_translate.py

  • Running the application:

    • Start By running the application backend from the api/api.py.
    • Run the application front from the mobile/main.py.

    Remark : The models are already trained and saved in models

  • Once the application is running you can choose between 3 models : digits, letters or cat-vs-dog model.

Scrapping :

• We implemented a scrapping functionnality in the project and to run it you have to install scrapping modules above and start dataset_generator from data/dataset_generator.py. You have to specifiy the number of pictures and what you are looking for.
• The result images are in the ressources directory devided into two sections : training and validation.

DataSets :

• If you want to use a valid datasets see the link below :

Cat VS Dog : https://github.com/abdellah-idris/catvsdog_dataset

letters : https://github.com/abdellah-idris/letters_dataset

digits : https://github.com/abdellah-idris/digit_data

###Remark : You need to verify the paths specified in the models and change them according to your dataset location.