GithubHelp home page GithubHelp logo

carnd-traffic_sign_classifier's Introduction

Traffic Sign Classifier

Udacity Self Driving Car Nanodegree project 2.

This project classify traffic sign with simple Convolutinon Network.

Dataset

German_Traffic_Sign_Dataset

I used the numpy and basic python operations to calculate summary statistics of the traffic signs data set.

  • The size of training set is 34799.
  • The size of test set is 12630.
  • The shape of a traffic sign image is (32, 32, 3).
  • The number of unique classes/labels in the data set is 43.

Preprocessing

Preprocessing pipeline has 3 steps.

  1. GrayScaling.
  2. Normalization with MinMaxScaler(scikit-learn).
  3. Shuffle data.

Model Architecture

Layer Description
Input 32x32x3 RGB image
Convolution 5x5 1x1 stride, same padding, outputs 32x32x8
RELU
Max pooling 2x2 stride, outputs 16x16x8
Convolution 5x5 1x1 stride, same padding, outputs 16x16x16
RELU
Max pooling 2x2 stride, outputs 8x8x16
Fully connected W.shap (8816, 300), dropout_rate(learn) 0.7
Fully connected W.shap (300, 120), dropout_rate(learn) 0.7
Fully connected W.shap (120, 43), dropout_rate(learn) 0.7
Softmax

Result

  • cost = 0.013
  • validation set accuracy = 0.949
  • test set accuracy = 0.942

Test

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

alt text alt text alt text alt text alt text

Here are the results of the prediction:

Image Prediction
No entry No entry
Stop Stop
Speed limit (60km/h) Speed limit (60km/h)
Go straight or left Go straight or left
Go straight or right Go straight or right

The model was able to correctly guess 5 of the 5 traffic signs, which gives an accuracy of 100%.

top-5 test output


alt text

Probability Prediction
1 No entry
1.1e-25 Stop
2.5e-28 Yield
5.9e-31 End of all speed and passing limits
2.8e-33 Speed limit (120km/h)

alt text

Probability Prediction
0.84 Stop
0.13 Yield
7.6e-3 No vehicles
4.4e-5 Speed limit (50km/h)
2.1e-6 Speed limit (60km/h)

alt text

Probability Prediction
0.99 Speed limit (60km/h)
1.0e-5 Speed limit (80km/h)
2.2e-7 End of speed limit (80km/h)
2.0e-8 Speed limit (50km/h)
2.4e-9 Speed limit (30km/h)

alt text

Probability Prediction
0.98 Go straight or left
1.4e-2 No entry
1.3e-5 Keep left
6.8e-8 Roundabout mandatory
6.8e-12 Ahead only

alt text

Probability Prediction
0.80 Go straight or right
0.15 No entry
0.05 Keep right
8.5e-4 End of all speed and passing limits
7.6e-4 End of no passing

carnd-traffic_sign_classifier's People

Contributors

dhkim0225 avatar

Watchers

 avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.