Simple image recognition using python with CIFAR10 dataset in Keras
I have used CIFAR10 dataset for training the model and size of the image is 32*32, You can alter it as you want and also use different activation functions.
Change the number of Epochs for more accuracy.
This is no complex Stuff, It's simple and easy to understand and also does not perform complex operations.
Steps for running:
- Use Pycharm or any IDE you like.
- Download Cifar dataset, if not downloaded or available in keras library.
- Set epoch in the trainer file as per your wish and for more accuracy.
- Only run tester, after training is done.
- Input the file path and you'll get the output.
P.S- You can also use CIFAR100 for more categories and better results.