Traffic Sign Classifier

This project was done as part of the Udacity Self-Driving Car Nanodegree

---

The goals / steps of this project are the following:

• Load the data set from INI
• Explore, summarize and visualize the data set
• Design, train and test a model architecture
• Use the model to make predictions on new images
• Analyze the softmax probabilities of the new images
• Summarize the results with a written report

---

Data Set Summary & Exploration

1. Provide a basic summary of the data set. In the code, the analysis should be done using python, numpy and/or pandas methods rather than hardcoding results manually.

I used the pandas library to calculate summary statistics of the traffic signs data set:

• The size of training set is 34799
• The size of the validation set is 4410
• The size of test set is 12630
• The shape of a traffic sign image is 32x32
• The number of unique classes/labels in the data set is 43

2. An exploratory visualization of the dataset.

Here is an exploratory visualization of the data set. It is a bar chart showing how the data is distributed in the training set.

Design and Test the Model Architecture

1. How the image data was preprocessed

As a first step, to pre-process the data, I decided to converted the images to grayscale. The shapes of the traffic signs are more important than the color and hence the model will train better. Next, the data was normalized between 0.1 and 0.9, which prevents overfitting.

I attempted to generate additional data as seen in the commented code in cell 7, but it resulted in less accurate results when testing with the images found on the internet.

Here is an example of an original image:

And here is the augmented image:

2. The final model architecture

My final model consisted of the following layers:

Layer	Description
Input	32x32x3 RGB image
Convolution 3x3	1x1 stride, same padding, outputs 32x32x64
RELU
Max pooling	2x2 stride, outputs 16x16x64
Convolution 3x3	1x1 stride, same padding, outputs 32x32x64
RELU
Max pooling	2x2 stride, outputs 16x16x64
Fully Connected	Outputs
RELU
DROPOUT	0.50
Fully Connected	Output
RELU
DROPOUT	0.50
Fully connected	Output
Softmax

3. How the model was trained

To train the model, I used an Adam Optimizer was used to try to minimize the loss. The learning rate was set to 0.001, Epochs to 50, and batch size of 128. There was two dropouts set to 0.50. The model was trained using TensorFlow and iterated over the Epochs, and the accuracy of which was measured.

4. The approach taken for finding the solution

My final model results were:

• training set accuracy of 0.999
• validation set accuracy of 0.966
• test set accuracy of 0.954

If a well known architecture was chosen:

• What architecture was chosen?
• Why did you believe it would be relevant to the traffic sign application?
• How does the final model's accuracy on the training, validation and test set provide evidence that the model is working well?

The LeNet architecture was chosen for this project as it is well-known and well-understood architecture. It is well-suited for identifying visual patterns and hence was ideal to the traffic sign application. The final model's accuracy of 0.999 for the training set, 0.966 for the validation set, and 0.954 for the test set, prove that the model works well for this application.

Test the Model on New Images

1. Five traffic signs found on the web to test the model

Here are five German traffic signs that I found on the web:

The fourth image might be difficult to classify because the shape is similiar to some other traffic signs.

2. Model's predictions on these traffic signs

Here are the results of the prediction:

Image	Prediction
Speed Limit: 70	Speed Limit: 70
No Passing	No Passing
Priority Road	Priority Road
Bumpy Road	Priority Road
Pedestrians	Pedestrians

The model was able to correctly guess 4 of the 5 traffic signs, which gives an accuracy of 80%.

3. The top 5 softmax probabilities for each image along with the sign type of each probability

The code for making predictions on my final model is located in the 11th cell of the Ipython notebook.

For the first image, the model is nearly certain that this is a stop sign (probability of 0.999993801), and the image does contain a speed limit sign. The top five soft max probabilities were:

Probability	Prediction
0.999993801	Speed Limit: 70
1.000000000	No Passing
0.999984026	Priority Road
0.65799582	Priority Road
0.852244556	Pedestrians

For the fourth image, the probability is notably lower than the others and is the only wrong predication.