new_load.py

from __future__ import division, print_function, absolute_import
import numpy as np
import tflearn
from tflearn.data_utils import shuffle, to_categorical
from tflearn.layers.core import input_data, dropout, fully_connected
from tflearn.layers.conv import conv_2d, max_pool_2d
from tflearn.layers.estimator import regression
from tflearn.data_preprocessing import ImagePreprocessing
from tflearn.data_augmentation import ImageAugmentation
import pickle

# Data loading and preprocessing

def my_model():
	
	input_layer = input_data(shape=[None, 50, 50, 1],name="input_layer")

	conv_1 = conv_2d(input_layer, 30, 3, activation='relu',name="conv_1")
	pool_1 = max_pool_2d(conv_1, 2,name="pool_1")

	conv_2 = conv_2d(pool_1, 30, 3, activation='relu',name="conv_2")
	pool_2 = max_pool_2d(conv_2, 2,name="pool_2")

	conv_3 = conv_2d(pool_2, 40, 3, activation='relu',name="conv_3")
	pool_3 = max_pool_2d(conv_3, 2,name="pool_3")

	conv_4 = conv_2d(pool_3, 40, 3, activation='relu',name="conv_4")
	pool_4 = max_pool_2d(conv_4, 2,name="pool_4")

	conv_5 = conv_2d(pool_4, 40, 3, activation='relu',name="conv_5")
	pool_5 = max_pool_2d(conv_5, 2,name="pool_5")

	conv_6 = conv_2d(pool_5, 30, 3, activation='relu',name="conv_6")
	pool_6 = max_pool_2d(conv_6, 2,name="pool_6")

	fc_1 = fully_connected(pool_6, 100, activation='relu',name="fc_1")
	drop = dropout(fc_1, 0.5,name="drop")

	fc_2 = fully_connected(drop, 50, activation='relu', name="fc_2")

	output = fully_connected(fc_2, 2, activation='softmax',name="output")
	network = regression(output)
	model = tflearn.DNN(network)
 	return model

# Train using classifier