sketch-recognition-ml-django

The Quick Draw Dataset is a collection of 50 million drawings across 345 categories, contributed by players of the game Quick, Draw!. The drawings were captured as timestamped vectors, tagged with metadata including what the player was asked to draw and in which country the player was located. You can browse the recognized drawings on quickdraw.withgoogle.com/data.

Dataset: quickdraw dataset

Note: Trained for only 10 labels class_names = { 0: 'ant', 1: 'bird', 2: 'crown', 3: 'envelope', 4: 'face', 5: 'fish', 6: 'flower', 7: 'house', 8: 'star', 9: 'sun' }

Code Requirements 🦄

You can install Conda, django and django REST for python which resolves all the dependencies for machine learning. or,

pip install -r requirements.txt

Implementation

1. Convolutional Neural Network
2. Django and Django REST framework powered backend

Setup 🖥️

1. Clone the repo
2. Install the pre requities libraries
3. On directory containing manage.py run

>>> python manage.py makemigrations

>>> python manage.py migrate

4. Create super user by running

>>> python manage.py createsuperuser

provide the details.

5. Run >>> python manage.py runserver
6. Visit http://127.0.0.1:8000/admin provide credentials for super user
7. Populate Label table with relavant details about the trained labels.(.svg images)
8. Visit http://127.0.0.1:8000/

---

Raised issues are greatly appreciated.

©️ Avishka Dambawinna __@2zunu