https://github.com/jdleesmiller/carnd-cloning

Train a driverless car to drive around a simulated race track using end-to-end deep learning — from camera images to steering commands. This is the code for my 'Zero to Driverless Race Car with Deep Learning' blog post and talk.
https://github.com/jdleesmiller/carnd-cloning

car convolutional-neural-networks deep-learning deep-neural-networks driverless inception ipython-notebook

Last synced: about 1 year ago
JSON representation

Train a driverless car to drive around a simulated race track using end-to-end deep learning — from camera images to steering commands. This is the code for my 'Zero to Driverless Race Car with Deep Learning' blog post and talk.

• Host: GitHub
• URL: https://github.com/jdleesmiller/carnd-cloning
• Owner: jdleesmiller
• License: mit
• Created: 2016-12-08T21:31:07.000Z (over 9 years ago)
• Default Branch: master
• Last Pushed: 2019-01-01T19:23:46.000Z (over 7 years ago)
• Last Synced: 2025-03-30T03:41:12.285Z (about 1 year ago)
• Topics: car, convolutional-neural-networks, deep-learning, deep-neural-networks, driverless, inception, ipython-notebook
• Language: Python
• Homepage: https://jdlm.info/articles/2019/01/01/driverless-race-car-deep-learning.html
• Size: 54.3 MB
• Stars: 5
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.ipynb
  • License: LICENSE

Awesome Lists containing this project

README

          
{

 "cells": [

  {

   "cell_type": "markdown",

   "metadata": {},

   "source": [

    "# README\n",

    "\n",

    "## Overall Approach\n",

    "\n",

    "- Take the image preprocessing and the first twelve layers (just convolutions) from the InceptionV3 network --- the `base_model`.\n",

    "- Apply a 3-sigma Gaussian filter to the steering angles in the training data to smooth them, because my keyboard input was quite spikey.\n",

    "- Also use the udacity training dataset, which was released after I collected my own data (but do not apply smoothing to it).\n",

    "- Run the images through the base model to produce bottleneck features.\n",

    "  - Also use the left and right camera images and bias the steering angle by 1.5 degrees toward the center.\n",

    "  - Also reflect the images horizontally and flip the sign of the angle.\n",

    "  - That gives 6 features per frame in total.\n",

    "- Train a deep neural net (see architecture below) with the bottleneck features as inputs.\n",

    "- Evaluate the model for roughly 100 different sets of hyperparameters (see Appendix) to produce a shortlist with low validation set Mean Absolute Error (MAE).\n",

    "- Evaluate the shortlist on the simulator.\n",

    "\n",

    "## Preliminaries"

   ]

  },

  {

   "cell_type": "code",

   "execution_count": 1,

   "metadata": {},

   "outputs": [],

   "source": [

    "%load_ext autoreload\n",

    "%autoreload 2\n",

    "%matplotlib inline"

   ]

  },

  {

   "cell_type": "code",

   "execution_count": 2,

   "metadata": {},

   "outputs": [

    {

     "name": "stderr",

     "output_type": "stream",

     "text": [

      "Using TensorFlow backend.\n",

      "/home/john/anaconda3/lib/python3.5/site-packages/sklearn/utils/fixes.py:313: FutureWarning: numpy not_equal will not check object identity in the future. The comparison did not return the same result as suggested by the identity (`is`)) and will change.\n",

      "  _nan_object_mask = _nan_object_array != _nan_object_array\n"

     ]

    }

   ],

   "source": [

    "import gzip\n",

    "import os\n",

    "import pickle\n",

    "import platform\n",

    "\n",

    "import cv2\n",

    "import keras\n",

    "import math\n",

    "import matplotlib.pyplot as plt\n",

    "import numpy as np\n",

    "import pandas as pd\n",

    "from scipy.misc import imread, imresize\n",

    "\n",

    "from common import * \n",

    "import bottleneck_features\n",

    "import preprocess\n",

    "import model_io\n",

    "import batch\n",

    "import model"

   ]

  },

  {

   "cell_type": "markdown",

   "metadata": {},

   "source": [

    "# Base Model\n",

    "\n",

    "The full inception model is too slow for real time use on my laptop, so I am just using the first few layers. Loading the whole inception model is also very slow, so cut out only the weights we need and save them for loading later."

   ]

  },

  {

   "cell_type": "code",

   "execution_count": 3,

   "metadata": {},

   "outputs": [],

   "source": [

    "import inception\n",

    "import model_io\n",

    "\n",

    "CUT_INDEX = 12 # layer 12 is before the first inception module\n",

    "\n",

    "if not os.path.exists(base_model_stem(CUT_INDEX) + '.json'):\n",

    "    model_io.save_model(\n",

    "        base_model_stem(CUT_INDEX) + '.json',\n",

    "        base_model_stem(CUT_INDEX) + '.h5',\n",

    "        inception.make_cut_model(CUT_INDEX))"

   ]

  },

  {

   "cell_type": "code",

   "execution_count": 4,

   "metadata": {},

   "outputs": [

    {

     "name": "stdout",

     "output_type": "stream",

     "text": [

      "____________________________________________________________________________________________________\n",

      "Layer (type)                     Output Shape          Param #     Connected to                     \n",

      "====================================================================================================\n",

      "input_1 (InputLayer)             (None, 160, 320, 3)   0                                            \n",

      "____________________________________________________________________________________________________\n",

      "convolution2d_3 (Convolution2D)  (None, 79, 159, 32)   896         input_1[0][0]                    \n",

      "____________________________________________________________________________________________________\n",

      "batchnormalization_1 (BatchNormal(None, 79, 159, 32)   64          convolution2d_3[0][0]            \n",

      "____________________________________________________________________________________________________\n",

      "convolution2d_4 (Convolution2D)  (None, 77, 157, 32)   9248        batchnormalization_1[0][0]       \n",

      "____________________________________________________________________________________________________\n",

      "batchnormalization_2 (BatchNormal(None, 77, 157, 32)   64          convolution2d_4[0][0]            \n",

      "____________________________________________________________________________________________________\n",

      "convolution2d_5 (Convolution2D)  (None, 77, 157, 64)   18496       batchnormalization_2[0][0]       \n",

      "____________________________________________________________________________________________________\n",

      "batchnormalization_3 (BatchNormal(None, 77, 157, 64)   128         convolution2d_5[0][0]            \n",

      "____________________________________________________________________________________________________\n",

      "maxpooling2d_1 (MaxPooling2D)    (None, 38, 78, 64)    0           batchnormalization_3[0][0]       \n",

      "____________________________________________________________________________________________________\n",

      "convolution2d_6 (Convolution2D)  (None, 38, 78, 80)    5200        maxpooling2d_1[0][0]             \n",

      "____________________________________________________________________________________________________\n",

      "batchnormalization_4 (BatchNormal(None, 38, 78, 80)    160         convolution2d_6[0][0]            \n",

      "____________________________________________________________________________________________________\n",

      "convolution2d_7 (Convolution2D)  (None, 36, 76, 192)   138432      batchnormalization_4[0][0]       \n",

      "____________________________________________________________________________________________________\n",

      "batchnormalization_5 (BatchNormal(None, 36, 76, 192)   384         convolution2d_7[0][0]            \n",

      "____________________________________________________________________________________________________\n",

      "maxpooling2d_2 (MaxPooling2D)    (None, 17, 37, 192)   0           batchnormalization_5[0][0]       \n",

      "====================================================================================================\n",

      "Total params: 173072\n",

      "____________________________________________________________________________________________________\n"

     ]

    }

   ],

   "source": [

    "model_io.load_base_model(CUT_INDEX).summary()"

   ]

  },

  {

   "cell_type": "markdown",

   "metadata": {},

   "source": [

    "## Preprocessing\n",

    "\n",

    "- Correct the image file paths, since I have several driving logs in different folders.\n",

    "- Smooth the control inputs. The input from me is quite spikey, because all I can do is press a button or not. This seems like it would make learning quite difficult, so apply smoothing.\n",

    "- Bottleneck the features to make training reasonably fast.\n",

    "\n",

    "### Training (and Validation) Data\n",

    "\n",

    "I started out by driving around quite fast and cutting corners (`ccw_1` and `cw_1`) and then doing several laps with recovery, as suggested in the assignment spec. I did both clockwise and counterclockwise on the first training track.\n",

    "\n",

    "Model performance was not very good on the fast data, so I recorded some slower (roughly 10mph) data, trying to be very careful about staying in the middle of the road. Performance was much better on the slower datasets. (I later modified `drive.py` to keep the speed at about 10mph to match the training data.)\n",

    "\n",

    "The main place where the model had problems was with the turnout after the bridge. It would often drive into it, and I did not provide any training data in there, so it usually got lost and crashed. I recorded four more drives around that corner, and now at least some of the models turn correctly to avoid the turnout."

   ]

  },

  {

   "cell_type": "code",

   "execution_count": 5,

   "metadata": {

    "scrolled": false

   },

   "outputs": [

    {

     "name": "stdout",

     "output_type": "stream",

     "text": [

      "data/ccw_2/bottleneck_12 exists; just binding.\n",

      "data/cw_2/bottleneck_12 exists; just binding.\n",

      "data/ccw_recover_from_right_1/bottleneck_12 exists; just binding.\n",

      "data/ccw_recover_from_left_1/bottleneck_12 exists; just binding.\n",

      "data/cw_recover_from_right_1/bottleneck_12 exists; just binding.\n",

      "data/cw_recover_from_left_1/bottleneck_12 exists; just binding.\n",

      "data/ccw_turnout_1/bottleneck_12 exists; just binding.\n",

      "data/ccw_turnout_2/bottleneck_12 exists; just binding.\n",

      "data/udacity/bottleneck_12 exists; just binding.\n",

      "                 bottleneck_features  brake  \\\n",

      "0  data/ccw_2/bottleneck_12/0000.npz    0.0   \n",

      "1  data/ccw_2/bottleneck_12/0001.npz    0.0   \n",

      "2  data/ccw_2/bottleneck_12/0002.npz    0.0   \n",

      "3  data/ccw_2/bottleneck_12/0003.npz    0.0   \n",

      "4  data/ccw_2/bottleneck_12/0004.npz    0.0   \n",

      "\n",

      "                                        center_image dataset  \\\n",

      "0  data/ccw_2/IMG/center_2016_12_10_17_02_46_943.jpg   ccw_2   \n",

      "1  data/ccw_2/IMG/center_2016_12_10_17_02_47_043.jpg   ccw_2   \n",

      "2  data/ccw_2/IMG/center_2016_12_10_17_02_47_143.jpg   ccw_2   \n",

      "3  data/ccw_2/IMG/center_2016_12_10_17_02_47_259.jpg   ccw_2   \n",

      "4  data/ccw_2/IMG/center_2016_12_10_17_02_47_360.jpg   ccw_2   \n",

      "\n",

      "                                        left_image  \\\n",

      "0  data/ccw_2/IMG/left_2016_12_10_17_02_46_943.jpg   \n",

      "1  data/ccw_2/IMG/left_2016_12_10_17_02_47_043.jpg   \n",

      "2  data/ccw_2/IMG/left_2016_12_10_17_02_47_143.jpg   \n",

      "3  data/ccw_2/IMG/left_2016_12_10_17_02_47_259.jpg   \n",

      "4  data/ccw_2/IMG/left_2016_12_10_17_02_47_360.jpg   \n",

      "\n",

      "                                        right_image  smooth_brake_1  \\\n",

      "0  data/ccw_2/IMG/right_2016_12_10_17_02_46_943.jpg             0.0   \n",

      "1  data/ccw_2/IMG/right_2016_12_10_17_02_47_043.jpg             0.0   \n",

      "2  data/ccw_2/IMG/right_2016_12_10_17_02_47_143.jpg             0.0   \n",

      "3  data/ccw_2/IMG/right_2016_12_10_17_02_47_259.jpg             0.0   \n",

      "4  data/ccw_2/IMG/right_2016_12_10_17_02_47_360.jpg             0.0   \n",

      "\n",

      "   smooth_brake_gaussian_3  smooth_brake_gaussian_5  smooth_steering_angle_1  \\\n",

      "0                      0.0                      0.0                -0.030136   \n",

      "1                      0.0                      0.0                -0.033305   \n",

      "2                      0.0                      0.0                -0.036808   \n",

      "3                      0.0                      0.0                -0.041335   \n",

      "4                      0.0                      0.0                -0.045728   \n",

      "\n",

      "   smooth_steering_angle_gaussian_3  smooth_steering_angle_gaussian_5  \\\n",

      "0                         -0.000397                         -0.010688   \n",

      "1                         -0.000872                         -0.012335   \n",

      "2                         -0.002138                         -0.015646   \n",

      "3                         -0.004856                         -0.020634   \n",

      "4                         -0.009930                         -0.027281   \n",

      "\n",

      "   smooth_throttle_1  smooth_throttle_gaussian_3  smooth_throttle_gaussian_5  \\\n",

      "0           0.124718                    0.004798                    0.080049   \n",

      "1           0.137835                    0.009238                    0.091355   \n",

      "2           0.152331                    0.020215                    0.113850   \n",

      "3           0.171067                    0.041763                    0.147185   \n",

      "4           0.189248                    0.079225                    0.190588   \n",

      "\n",

      "      speed  steering_angle  throttle                    time  \n",

      "0  0.000079             0.0       0.0 2016-12-10 17:02:46.943  \n",

      "1  0.000079             0.0       0.0 2016-12-10 17:02:47.043  \n",

      "2  0.000078             0.0       0.0 2016-12-10 17:02:47.143  \n",

      "3  0.000078             0.0       0.0 2016-12-10 17:02:47.259  \n",

      "4  0.000078             0.0       0.0 2016-12-10 17:02:47.360  \n"

     ]

    }

   ],

   "source": [

    "log = pd.concat([\n",

    "    # preprocess.run('data/ccw_1', CUT_INDEX), # driving fairly aggressively\n",

    "    # preprocess.run('data/cw_1', CUT_INDEX),\n",

    "    preprocess.run('data/ccw_2', CUT_INDEX),\n",

    "    preprocess.run('data/cw_2', CUT_INDEX),\n",

    "    preprocess.run('data/ccw_recover_from_right_1', CUT_INDEX),\n",

    "    preprocess.run('data/ccw_recover_from_left_1', CUT_INDEX),\n",

    "    preprocess.run('data/cw_recover_from_right_1', CUT_INDEX),\n",

    "    preprocess.run('data/cw_recover_from_left_1', CUT_INDEX),\n",

    "    preprocess.run('data/ccw_turnout_1', CUT_INDEX),\n",

    "    preprocess.run('data/ccw_turnout_2', CUT_INDEX),\n",

    "    preprocess.run('data/udacity', CUT_INDEX, header=0, smooth=False)\n",

    "], sort=True)\n",

    "print(log.head())"

   ]

  },

  {

   "cell_type": "markdown",

   "metadata": {},

   "source": [

    "###  Dataset Summary\n"

   ]

  },

  {

   "cell_type": "code",

   "execution_count": 6,

   "metadata": {

    "scrolled": true

   },

   "outputs": [

    {

     "data": {

      "text/html": [

       "
\n",

       "\n",

       "    .dataframe tbody tr th:only-of-type {\n",

       "        vertical-align: middle;\n",

       "    }\n",

       "\n",

       "    .dataframe tbody tr th {\n",

       "        vertical-align: top;\n",

       "    }\n",

       "\n",

       "    .dataframe thead th {\n",

       "        text-align: right;\n",

       "    }\n",

       "\n",

       "\n",

       "  \n",

       "    \n",

       "      \n",

       "      len\n",

       "      speed\n",

       "      steering_angle\n",

       "    \n",

       "    \n",

       "      dataset\n",

       "      \n",

       "      \n",

       "      \n",

       "    \n",

       "  \n",

       "  \n",

       "    \n",

       "      ccw_2\n",

       "      2882\n",

       "      8.833225\n",

       "      -0.041838\n",

       "    \n",

       "    \n",

       "      ccw_recover_from_left_1\n",

       "      1405\n",

       "      11.775736\n",

       "      0.050748\n",

       "    \n",

       "    \n",

       "      ccw_recover_from_right_1\n",

       "      1361\n",

       "      10.212416\n",

       "      -0.123761\n",

       "    \n",

       "    \n",

       "      ccw_turnout_1\n",

       "      713\n",

       "      9.905384\n",

       "      -0.085719\n",

       "    \n",

       "    \n",

       "      ccw_turnout_2\n",

       "      522\n",

       "      9.295456\n",

       "      -0.111309\n",

       "    \n",

       "    \n",

       "      cw_2\n",

       "      2391\n",

       "      9.972545\n",

       "      0.036800\n",

       "    \n",

       "    \n",

       "      cw_recover_from_left_1\n",

       "      1550\n",

       "      10.484323\n",

       "      0.128920\n",

       "    \n",

       "    \n",

       "      cw_recover_from_right_1\n",

       "      1721\n",

       "      8.972610\n",

       "      0.017962\n",

       "    \n",

       "    \n",

       "      udacity\n",

       "      8036\n",

       "      28.169839\n",

       "      0.004070\n",

       "    \n",

       "  \n",

       "\n",

       "
"

      ],

      "text/plain": [

       "                           len      speed  steering_angle\n",

       "dataset                                                  \n",

       "ccw_2                     2882   8.833225       -0.041838\n",

       "ccw_recover_from_left_1   1405  11.775736        0.050748\n",

       "ccw_recover_from_right_1  1361  10.212416       -0.123761\n",

       "ccw_turnout_1              713   9.905384       -0.085719\n",

       "ccw_turnout_2              522   9.295456       -0.111309\n",

       "cw_2                      2391   9.972545        0.036800\n",

       "cw_recover_from_left_1    1550  10.484323        0.128920\n",

       "cw_recover_from_right_1   1721   8.972610        0.017962\n",

       "udacity                   8036  28.169839        0.004070"

      ]

     },

     "execution_count": 6,

     "metadata": {},

     "output_type": "execute_result"

    }

   ],

   "source": [

    "def summarize_dataset():\n",

    "    groups = log.groupby('dataset')\n",

    "    totals = groups['center_image'].agg([len])\n",

    "    mean_speed = groups['speed'].mean()\n",

    "    mean_steering_angle = groups['steering_angle'].mean()\n",

    "    return pd.concat([totals, mean_speed, mean_steering_angle], axis=1)\n",

    "summarize_dataset()"

   ]

  },

  {

   "cell_type": "markdown",

   "metadata": {},

   "source": [

    "### Example Images with Bottleneck Features\n",

    "\n",

    "Here are example images from a single frame. Images from each of the three cameras are included, together with their horizontal reflections. The first three feature channels in the bottleneck features for each image are also plotted."

   ]

  },

  {

   "cell_type": "code",

   "execution_count": 7,

   "metadata": {

    "scrolled": false

   },

   "outputs": [

    {

     "data": {

      "image/png": "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