Can I train using my own dataset, for face landmarks detection using SAN my dataset totally 20,000 face image. It is enough? If it is possible what steps I have to do for it
In my case I want try SAN for face landmarks detection but with my own dataset. I could not understand the step for it
Enviroment:
Windows 10
Python 3.6
Can I train using my own dataset, for face landmarks detection using SAN my dataset totally 20,000 face image. It is enough? If it is possible what steps I have to do for it