Convolutional Neural Network to Classify Dogs and Cat. I built an ImageClassifier which classifies and tells you whether its a Dog image or a Cat image. I built a convolutional network which consists of Three Convolution layer and Three MaxPooling layer. Each Convolutional layer has filters, kernel size. Maxpooling layer has stride and pooling size. Then this Convolutional layer Connects to DeepNeuralNetwork. DNN has three hidden layer and output layer having Sigmoid Activation function. I trained this model for 31 epochs and achieved an accuracy of around 85%(Training Time=40 minutes using NVIDIA P100 GPU). I found this massive image dataset online which has 10,028 images(Ten Thousand and Twenty Eight). My model Predicted accurately during the testing phase. I even tested my model using my neighbor dog's pic and it predicted accurately.
I could have improved my model's accuracy even more by further increasing the number of epochs which would have taken a lot more training time.