Author: lexfridman

tutorial_driving_scene_segmentation/tutorial_driving_scene_segmentation.ipynb

@@ -87,7 +87,9 @@
    "metadata": {},
    "source": [
     "### Build the model\n",
+    "\n",
     "**[DeepLab](https://github.com/tensorflow/models/tree/master/research/deeplab)** is a state-of-art deep learning model for semantic image segmentation, where the goal is to assign semantic labels (e.g., person, dog, cat and so on) to every pixel in the input image. Some segmentation results on Flickr images:\n",
+    "\n",
     "<p align=\"center\">\n",
     "    <img src=\"https://github.com/tensorflow/models/blob/master/research/deeplab/g3doc/img/vis1.png?raw=true\" width=600></br>\n",
     "    <img src=\"https://github.com/tensorflow/models/blob/master/research/deeplab/g3doc/img/vis2.png?raw=true\" width=600></br>\n",
@@ -532,16 +534,29 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "evaluating on image MIT_DriveSeg_sample.png...\n",
+      "evaluating on image mit_driveseg_sample.png...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:19: DeprecationWarning: The binary mode of fromstring is deprecated, as it behaves surprisingly on unicode inputs. Use frombuffer instead\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
       "pixel accuracy: 0.90720\n",
-      "mean class IoU: 0.538374\n",
+      "mean class IoU: 0.5383739999999999\n",
       "class IoU:\n",
       "   road    sidewalk    building     pole    traffic light    traffic sign    vegetation    terrain    person     car\n",
       "-------  ----------  ----------  -------  ---------------  --------------  ------------  ---------  --------  ------\n",