HackPSU_Self_Driving_Car

• 

  Loss function of Train vs Validation

• 

  Test vs TRain accuracy

• 

  Supersonic Sensor

• 

  Raspberry Pi, Servo Motor, Power board

• 

  Gpio connections

(Train & Test) plot against Epochs (iterations)

Plot to find the optimum number of Epochs (iterations) for Artificial Neural Network (for the setting of the car using Camera resolution 320x240 pixels, Batch Size = 40, Learnig Rate =0.0025,) 1) Accuracy(Train & Test) plot against Epochs(iterations) 2) Loss Function (Train & Test) plot against Epochs (iterations)

The files we upload consists of

1)The Neural Network code to fit the model to be used in driving the car (All the dataset are not shown here, but you can directly pm me if needed)

2)The validation code (Train&Test) for Neural Network to test for the accuracy (the model achieved 90.21% accuracy for Testing Data, and 90.21% accuracy for Training Data)

3)The Feed Forward Neural Network to make the decision while running the car

4) Cleaning Code

5) graph plotting code

6) Getting the NN Parameter code from the model (extract from the file .h5)

7) Code to run the car autonomously

8) Code to train the car

Built With

• python