- [Cats and Dogs](#Cats and Dogs)
Implementation a Binary Classifier to recognize pets (cats or dogs) in an image. Check out my kernel here The model is a trimmed version of VGG16 with 1 dilated convolution layer and a single FC layer that uses swish activation function.
Trainable params: 360,865
|
|
Distribution of Image Height-to-Width Ratio for cats and dogs class
|
|
-
Training: After 8 Epochs ==> loss: 0.2833 - acc: 0.8794 - val_loss: 0.4017 - val_acc: 0.8215
-
Prediction
|
|
|
|
Contact: Let me know if you have questions/suggestions. Feel free to connect with me on LinkedIn.