Flower_Recognition_CNN

Project description

a CNN-based flower classifier in Keras

Data

Right now, the dataset contains 6 types of flowers from ImageNet which are cherry, chrysanthemum, Jasmine, narcissus, Rose, Sun_flower.

Training

The training dataset has 540 samples for each kind of flower.

The test dataset has 130 samples for each kind of flower.

Evaluation

The classifier achieves over 95% accuracy on validation set but on test set, only 75%, which needs to handle later...