• finish design
• use react
• start frontend first

• fix decision on using openCV
• split image manipulation to sections
• face section -> detect face and add accessories with openCV
• might add CNN to predict emotions, age and sex
• image filter -> invert color, edge detection with openCV

• image manipulation using DCGAN is computationally expensive
• might take days to train the model without GPU
• hardware is not strong enough to handle that computations
• decided to learn and use openCV instead
• cartoonize, line detection, object recognition

Plan:

• User can upload image and choose what to do with the image (face recognition, edge detection, etc)
• Image is sent to the server and the progress is shown in the frontend
• Server starts asynchronous task
• The server use a deep convolutional generative adversarial networks (DCGAN) model to process the image
• while also sending progress to the frontend
• When image manipulation is done, image is sent back from the server to the frontend

Tools:

• Django -> backend server
• django rest framework -> build api
• DCGAN -> tensorflow / pytorch / opencv / sklearn (?, unsure)
• React / Vue -> frontend
• axios -> frontend api requests
• Celery -> handle asynchronous tasks
• WebSocket -> send task progress to the frontend (?, dont know if needed)
• redis / RabbitMQ -> message broker for celery

Deployment:

• Heroku -> cloud service
• Channels -> serve asgi application
• nginx -> serve static resources

Topics:

• Backend development: sessions, message broker, async tasks
• Frontend development: responsive, user experience, web design
• Machine learning / Deep learning: unsupervised machine learning generate something