In this project, you will use what you've learned about deep neural networks and convolutional neural networks to classify traffic signs. You will train and validate a model so it can classify traffic sign images using the German Traffic Sign Dataset. After the model is trained, you will then try out your model on images of German traffic signs that you find on the web.
The goals / steps of this project are the following:
- Load the data set
- 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
This lab requires:
- CarND Term1 Starter Kit The lab environment can be created with CarND Term1 Starter Kit. Click here for the details.
- The code start from extracting traffice-sign-data set of three category
- traning
- validation
- test
- Piepline consist of 5 set of step's
- Reading image
- checking the size of all the images
- plotting the graph on Labels vs Classes occurance in program
- Grayscale the image in order to obtain one color channel
- optiization via LeNet architecture
- Lenet architecture take input of 32,32,1 input size image
- Having 5 convolution Layer
- Having process to get (weight , bias , convolution or matmul ),using Activation funtion such as relu , and pooling , sometimes drop out is also used for best results
- final output will be 84
- trainig pipe consist of LeNet architechture , softmax , training operation 1 traingin operation use AdamOptimizer for backpropogation and lossoperation such as mean of data
- evolution model is used
- combining above pipes we train model , run validation set on it to know the accuraty usign evalutin pipe , run test data set and get 93.2 % prediction
- The Contrast of the image.
- The Angle of the traffic sign.
- Image might be jittered.
- noramlization can be applited