Image Classification CNN

Train a model from a dataset for predicting cat and dog from image.

Training

Epoch 1/5
32/32 [==============================] - 7s 199ms/step - loss: 0.6948 - accuracy: 0.5310
Epoch 2/5
32/32 [==============================] - 6s 200ms/step - loss: 0.6518 - accuracy: 0.6170
Epoch 3/5
32/32 [==============================] - 7s 207ms/step - loss: 0.5855 - accuracy: 0.6830
Epoch 4/5
32/32 [==============================] - 7s 217ms/step - loss: 0.5096 - accuracy: 0.7505
Epoch 5/5
32/32 [==============================] - 7s 229ms/step - loss: 0.4456 - accuracy: 0.7940
13/13 [==============================] - 0s 26ms/step - loss: 0.6087 - accuracy: 0.6925

Using

$ python app.py

Demo CNN Model

Image1	Label1	Predict1	Image2	Label2	Predict2
	DOG	DOG		CAT	CAT
	CAT	DOG		CAT	CAT
	DOG	DOG

Sources

• https://drive.google.com/drive/u/0/folders/1dZvL1gi5QLwOGrfdn9XEsi4EnXx535bD
• https://keras.io/api/losses/