Behaviorial Cloning Project

The goals / steps of this project are the following:

• Use the simulator to collect data of good driving behavior
• Build, a convolution neural network in Keras that predicts steering angles from images
• Train and validate the model with a training and validation set
• Test that the model successfully drives around track one without leaving the road
• Summarize the results with a written report

Rubric Points

Here I will consider the rubric points individually and describe how I addressed each point in my implementation.

---

Files Submitted & Code Quality

1. Submission includes all required files and can be used to run the simulator in autonomous mode

My project includes the following files:

• model.py containing the script to create and train the model
• drive.py for driving the car in autonomous mode
• model.h5 containing a trained convolution neural network
• README.md summarizing the results

2. Submission includes functional code

Using the Udacity provided simulator and my drive.py file, the car can be driven autonomously around the track by executing

python drive.py model.h5

3. Submission code is usable and readable

The model.py file contains the code for training and saving the convolution neural network. The file shows the pipeline I used for training and validating the model, and it contains comments to explain how the code works.

Model Architecture and Training Strategy

1. An appropriate model architecture has been employed

My model consists of a convolution neural network with 5x5 filter sizes, 2x2 subsampling, and depths between 24 and 64 (model-gen.py lines 63-76)

The model includes RELU layers to introduce nonlinearity (code line 71, 73, and 75), and the data is normalized in the model using a Keras lambda layer (code line 63).

2. Attempts to reduce overfitting in the model

The model contains dropout layers in order to reduce overfitting (model-gen.py lines 72 and 74).

The model was trained and validated on different data sets to ensure that the model was not overfitting (code line 10-16). The model was tested by running it through the simulator and ensuring that the vehicle could stay on the track.

3. Model parameter tuning

The model used an adam optimizer, so the learning rate was not tuned manually (model.py line 78).

4. Appropriate training data

Training data will be covered in the section Creation of the Training Set and Training Process

Model Architecture and Training Strategy

1. Solution Design Approach

The overall strategy for deriving a model architecture was to ...

My first step was to use a convolution neural network model similar to the network in the Even More Powerful Network section. I thought this model might be appropriate because its initial results out the gate were strong and only needed more training data as opposed to more network layers.

As I began to have trouble with the validation accuracy and testing using drive.py I introduced more training data for tricky areas, particular edges for curves. After I determined my training data to be sufficient, I found the model to be overfitting around epoch 3. In response, I added dropout to the bottom 2 fully connected layers and found the model was better avoiding overfitting.

In order to better improve the the model taken from Even More Powerful Network I increased the fully connected layers number of parameters to 200, 100, 50, 1.

The final step was to run the simulator to see how well the car was driving around track one. There were a few spots where the vehicle fell off the track. To improve the driving behavior in these cases, I added more short clips of moving from the outside of a curve to the inside quickly and precisely.

At the end of the process, the vehicle is able to drive autonomously around the track without leaving the road.

2. Final Model Architecture

The final model architecture (model.py lines 65-76) consisted of a convolution neural network with the following layers and layer sizes.

model.add(Convolution2D(24,5,5,subsample=(2,2),activation='relu'))
model.add(Convolution2D(36,5,5,subsample=(2,2),activation='relu'))
model.add(Convolution2D(48,5,5,subsample=(2,2),activation='relu'))
model.add(Convolution2D(64,3,3,activation='relu'))
model.add(Convolution2D(64,3,3,activation='relu'))
model.add(Flatten())
model.add(Dense(200, activation='relu'))
model.add(Dropout(0.6))
model.add(Dense(100, activation='relu'))
model.add(Dropout(0.6))
model.add(Dense(50, activation='relu'))
model.add(Dense(1))

3. Creation of the Training Set & Training Process

Choosing the training data was certainly the most important aspect of this project. My training data consisted of the following:

• 2 optimal laps around the track maintaining centrality in the lane as much as possible
• Small clips starting at the edges of the track and moving to the middle. Emphasis was placed on the edges of curves as those were particularly problematic during inital tests.