Image  Classification

4 Processes we work on when we do image classification.

1. Convolution : Means for identification of features.

2. Pooling : Reduces Image  Size without losing features.

3. Flattening : Convert 2-D Array to 1-D Array.

4. Full Connection: Feed 1-D Array to Neural Network.

M/L Platforms

1. Caffe                                                                                                                                                 2.Torch

3.TensorFlow                                                                                                                                         4.Keras

Keras (M/L Platform) Commands ( For Learning Purpose)

  1. Model = Sequential()
  2. Model.add (Conditional 2D(Filter,Kernel-Size,Input Shape))
  3. Model.add(MaxPooling2D(pool_size),i)
  4. Model.add(Flatten())
  5. Model.add(Dense(Units=No,activation))
  6. Model.add(Dense(Units =1,Activation))
  7. Model.compile(optimizer,loss,metres)                                                                                                                                                                                                                                                                                                                                                                                             Imp Points                                                    1. Epochs                              2.Batch_Size                      3. Training Sets                 4.Test Sets                       5. Dropouts                           6.Activation Functions        7. Dropouts                       8. Pool-Size                      9. Loss                                  10. Optimizer                     11. Metrics                        12. Rescale