My training data had only 1500 images in different sizes,
so I generated further images by rotating and flipping and mirroring which was approx 5000 images.
And coverted all images to fixed size.
The thing was had only 99 species of leaves of images.
That it self was heavy for me.
I never knew I could do it.
My friend suggested to keep a dead line and do this.
Which did work out, I was wondering how its even possible, but I made it happen.
Got the data online.
Its fun actually when I do things like this.
In udemy there are very good courses is less price as there is alway seasonal sales going on.
Start from : [2021 Complete Python Bootcamp From Zero to Hero in Python] and I have reached till NLP which also looks easy.
The person who teach is really good.
My final goal is to make a chat bot.
Here is some thing interesting this year:
