Initial commit

Tensorflow/tutoriel30/train.py

@@ -0,0 +1,50 @@
+import tensorflow as tf
+import numpy as np
+import glob
+import cv2
+import model
+import config
+
+tab_images=[]
+tab_labels=[]
+
+def complete_dataset(files, value):
+    for image in glob.glob(files):
+        img=cv2.imread(image)
+        img=cv2.resize(img, (config.size, config.size))
+        tab_images.append(img)
+        tab_labels.append([value])
+        img=cv2.flip(img, 1)
+        tab_images.append(img)
+        tab_labels.append([value])
+        img=cv2.flip(img, 0)
+        tab_images.append(img)
+        tab_labels.append([value])
+
+complete_dataset(config.dir_pos+'\\*.png', 1.)
+complete_dataset(config.dir_neg+'\\*.png', 0.)
+
+tab_images=np.array(tab_images, dtype=np.float32)/255
+tab_labels=np.array(tab_labels, dtype=np.float32)
+
+index=np.random.permutation(len(tab_images))
+tab_images=tab_images[index]
+tab_labels=tab_labels[index]
+
+#for i in range(len(tab_images)):
+#    cv2.imshow('Camera', tab_images[i])
+#    print("Label", tab_labels[i])
+#    if cv2.waitKey()&0xFF==ord('q'):
+#           quit()
+
+model=model.model(config.size, 8)
+
+model.compile(optimizer='adam',
+              loss='binary_crossentropy',
+              metrics=['accuracy'])
+model.fit(tab_images,
+          tab_labels,
+          validation_split=0.05,
+          batch_size=64,
+          epochs=30)
+model.save('saved_model\\my_model')