Behavioral 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

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
• writeup_report.md or writeup_report.pdf summarizing the results

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

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.

First, the model 'cut' the sky and front of car, with the function Cropping2D after that, normalize the data and centering this.

Like I said before, I use the NVIDIA neural network, which contain 5 convolutional layer and 3 Dense. The first 3 convolutional layer use 5x5 filter and 2x2 strides and the other 2 use 3x3 filters and 1x1 stride. After this, flatten the data and put 3 Dense layer.

I use augmented data for reduce overfitting.

The model used an adam optimizer, so the learning rate was not tuned manually.

Training data was chosen to keep the vehicle driving on the road. I used a combination of center lane driving, recovering from the left and right sides of the road and take smooth and slow curve to learn better the stearing on the road.

I choose the NVIDIA model. I put the architecture and training for 7 epochs, the training and validation loss are almost the same (0.023, 0.024). The most important thing I can see was the amounnt of data that we need to train the model.

The unique change on the model was, the convolutional layers had valid padding, I used from the second layer same padding. With this configuration I have better results.

Here is a visualization of the architecture.

I drive for 3 laps in the center of the road, sometimes I go to the edges and recover the center. I drive the last lap slow for take smooth curve.

With this amount of data I add a little more with cv2.flip, with this is "like we drive counter clockwise".

The preprocessing step was made with lambda function.

The 80% of the data was for training, the rest was use for validation.# clonning_behaviour