Hi,
I'm trying to load a model that I trained in Keras with OpenCV Dnn model.
I converted the model into .pb and .pbtxt files following this post.
However the final model outputs doesn't make any sense. So after browsing other forums I'm still lost/confused about the steps that it is needed to follow to do this conversion. I think that I'm not converting it properly to .pb and .pbtxt files.
So here is my model, it has 5 classes and :
num_classes =
5 5
epochs =
50 50
img_x, img_y = 32,
15 15
input_shape
=
= (img_x, img_y,
3) 3)
X_train, X_test, y_train, y_test
=
= train_test_split(data_x,
labels,
labels, test_size=0.20, random_state=42)
X_train, X_val, y_train, y_val
=
= train_test_split(X_train,
y_train,
y_train, test_size=0.2,
random_state=42) random_state=42)
X_train = X_train.astype('float32')
X_test = X_test.astype('float32')
X_val = X_val.astype('float32')
X_train /=
255 255
X_test /=
255 255
X_val
/=
255 /= 255
print('x_train shape:', X_train.shape)
print(X_train.shape[0],
'train
samples') 'train samples')
print(X_test.shape[0],
'test
samples') 'test samples')
y_train
=
keras.utils.to_categorical(y_train,
num_classes) = keras.utils.to_categorical(y_train, num_classes)
y_test
=
keras.utils.to_categorical(y_test,
num_classes) = keras.utils.to_categorical(y_test, num_classes)
model = Sequential()
model.add(Conv2D(32,
kernel_size=(5,
kernel_size=(5, 5),
strides=(1,
1),activation='relu',input_shape=input_shape,
strides=(1, 1),activation='relu',input_shape=input_shape, padding='same'))
model.add(MaxPooling2D(pool_size=(2,
model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2)))
model.add(Conv2D(64, (5,
5),
5), activation='relu', padding='same'))
model.add(MaxPooling2D(pool_size=(2,
2))) model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Flatten())
model.add(Dense(1000,
model.add(Dense(1000, activation='relu'))
model.add(Dense(num_classes,
model.add(Dense(num_classes, activation='softmax', name
=
"output_node")) = "output_node"))
model.compile(loss=keras.losses.categorical_crossentropy,
optimizer=keras.optimizers.Adadelta(),
metrics=['accuracy']) metrics=['accuracy'])
Can someone tell me step buy step how do I properly convert it to the tensorflow files?
