Updated imageNet models notebook

Author: Anastasia

Keras-ImageNet-Models/images/pumpkins_output.jpg

@@ -57,7 +57,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# instantiate the models\n",
+    "# init the models\n",
     "vgg_model = vgg16.VGG16(weights='imagenet')\n",
     "\n",
     "inception_model = inception_v3.InceptionV3(weights='imagenet')\n",
@@ -71,7 +71,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Let's define the image for further experiments assigning its location to the ``filename`` variable. You can choose any other image example to test the models."
+    "Let's define the image path for further experiments in the ``filename`` variable. Please note, that there are other images in the ``images`` directory that you can use as well to test the models."
    ]
   },
   {
@@ -144,7 +144,7 @@
     {
      "data": {
       "text/plain": [
-       "<matplotlib.image.AxesImage at 0x7f23000a2510>"
+       "<matplotlib.image.AxesImage at 0x7f3270126c50>"
       ]
      },
      "execution_count": 5,
@@ -212,18 +212,15 @@
    "metadata": {},
    "outputs": [
     {
-     "data": {
-      "text/plain": [
-       "[[('n02123597', 'Siamese_cat', 0.30934104),\n",
-       "  ('n01877812', 'wallaby', 0.08034121),\n",
-       "  ('n02326432', 'hare', 0.07509851),\n",
-       "  ('n02325366', 'wood_rabbit', 0.05053079),\n",
-       "  ('n03223299', 'doormat', 0.048173636)]]"
-      ]
-     },
-     "execution_count": 6,
-     "metadata": {},
-     "output_type": "execute_result"
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "('n02123597', 'Siamese_cat', 0.30934194)\n",
+      "('n01877812', 'wallaby', 0.08034124)\n",
+      "('n02326432', 'hare', 0.07509843)\n",
+      "('n02325366', 'wood_rabbit', 0.0505307)\n",
+      "('n03223299', 'doormat', 0.048173614)\n"
+     ]
     }
    ],
    "source": [
@@ -234,9 +231,11 @@
     "predictions = vgg_model.predict(processed_image)\n",
     "# print predictions\n",
     "# convert the probabilities to class labels\n",
-    "# We will get top 5 predictions which is the default\n",
+    "# we will get top 5 predictions which is the default\n",
     "label_vgg = decode_predictions(predictions)\n",
-    "label_vgg"
+    "# print VGG16 predictions\n",
+    "for prediction_id in range(len(label_vgg[0])):\n",
+    "    print(label_vgg[0][prediction_id])"
    ]
   },
   {
@@ -252,16 +251,13 @@
    "metadata": {},
    "outputs": [
     {
-     "data": {
-      "text/plain": [
-       "[[('n02124075', 'Egyptian_cat', 0.15737584),\n",
-       "  ('n03958227', 'plastic_bag', 0.14362834),\n",
-       "  ('n03223299', 'doormat', 0.14099452)]]"
-      ]
-     },
-     "execution_count": 7,
-     "metadata": {},
-     "output_type": "execute_result"
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "('n02124075', 'Egyptian_cat', 0.15737602)\n",
+      "('n03958227', 'plastic_bag', 0.14362879)\n",
+      "('n03223299', 'doormat', 0.14099468)\n"
+     ]
     }
    ],
    "source": [
@@ -274,7 +270,9 @@
     "# convert the probabilities to class labels\n",
     "# If you want to see the top 3 predictions, specify it using the top argument\n",
     "label_resnet = decode_predictions(predictions, top=3)\n",
-    "label_resnet"
+    "# print ResNet predictions\n",
+    "for prediction_id in range(len(label_resnet[0])):\n",
+    "    print(label_resnet[0][prediction_id])"
    ]
   },
   {
@@ -290,18 +288,15 @@
    "metadata": {},
    "outputs": [
     {
-     "data": {
-      "text/plain": [
-       "[[('n03958227', 'plastic_bag', 0.15703735),\n",
-       "  ('n02124075', 'Egyptian_cat', 0.11697447),\n",
-       "  ('n02123597', 'Siamese_cat', 0.10532668),\n",
-       "  ('n02123045', 'tabby', 0.075648665),\n",
-       "  ('n02909870', 'bucket', 0.054680943)]]"
-      ]
-     },
-     "execution_count": 8,
-     "metadata": {},
-     "output_type": "execute_result"
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "('n03958227', 'plastic_bag', 0.1570367)\n",
+      "('n02124075', 'Egyptian_cat', 0.11697376)\n",
+      "('n02123597', 'Siamese_cat', 0.10532685)\n",
+      "('n02123045', 'tabby', 0.0756485)\n",
+      "('n02909870', 'bucket', 0.054681662)\n"
+     ]
     }
    ],
    "source": [
@@ -313,7 +308,9 @@
     "\n",
     "# convert the probabilities to imagenet class labels\n",
     "label_mobilenet = decode_predictions(predictions)\n",
-    "label_mobilenet"
+    "# print MobileNet predictions\n",
+    "for prediction_id in range(len(label_mobilenet[0])):\n",
+    "    print(label_mobilenet[0][prediction_id])"
    ]
   },
   {
@@ -332,18 +329,15 @@
    "metadata": {},
    "outputs": [
     {
-     "data": {
-      "text/plain": [
-       "[[('n02124075', 'Egyptian_cat', 0.66225773),\n",
-       "  ('n02123045', 'tabby', 0.050285283),\n",
-       "  ('n02123597', 'Siamese_cat', 0.03638168),\n",
-       "  ('n02123159', 'tiger_cat', 0.023522485),\n",
-       "  ('n03223299', 'doormat', 0.0152056785)]]"
-      ]
-     },
-     "execution_count": 9,
-     "metadata": {},
-     "output_type": "execute_result"
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "('n02124075', 'Egyptian_cat', 0.66225827)\n",
+      "('n02123045', 'tabby', 0.05028525)\n",
+      "('n02123597', 'Siamese_cat', 0.03638152)\n",
+      "('n02123159', 'tiger_cat', 0.023522492)\n",
+      "('n03223299', 'doormat', 0.015205677)\n"
+     ]
     }
    ],
    "source": [
@@ -364,7 +358,9 @@
     "\n",
     "# convert the probabilities to class labels\n",
     "label_inception = decode_predictions(predictions)\n",
-    "label_inception"
+    "# print Inception predictions\n",
+    "for prediction_id in range(len(label_inception[0])):\n",
+    "    print(label_inception[0][prediction_id])"
    ]
   },
   {
@@ -404,7 +400,7 @@
     {
      "data": {
       "text/plain": [
-       "(-0.5, 699.5, 699.5, -0.5)"
+       "<matplotlib.image.AxesImage at 0x7f32381f4a10>"
       ]
      },
      "execution_count": 11,
@@ -426,8 +422,8 @@
    ],
    "source": [
     "plt.figure(figsize=[10,10])\n",
-    "plt.imshow(numpy_image)\n",
-    "plt.axis('off')"
+    "plt.axis('off')\n",
+    "plt.imshow(numpy_image)"
    ]
   }
  ],