FishHunt-Backend

The python backend part of stationed kid's game made for Primorsky Aquarium. Writen in tensorflow and django rest framework

Modules

Prodject consists of

• Classificator (Neural model)
• Django REST Backend

Deploing

1. install requirmented librarys

   pip install -r requirments.txt
   

2. Upload a dataset for training. It must have a following structure (there may be any fish folders ofc):

   Dataset/
   ├─ Fish_Dataset/
   │  ├─ Clownfish/
   │  │  ├─ *.jpg
   │  ├─ Swordfish/
   │  │  ├─ *.jpg
   

3. Run a create_model.ipynb notebook to create your's neural model

   jupyter nbconvert --execute create_model.ipynb
   

4. Migrate a database

   python Fishhunt/manage.py migrate
   

5. Run a server

   python Fishhunt/manage.py runserver
   

Usage

1. Create user for django admin panel

   python manage.py createsuperuser
   

2. Got to localhost:8000/admin to create fish models

   You can get model_label parameters from labels var in notebook

License

This code is totally free to use and modify if you have concrete author

Thanks to

• All personel Department of pre-university education
• My frontend teammates
• Andrey and Matvey: you are weak teachers but greate mates <3