traffic-sign-classifier

Used Convolutional neural networks (CNN) for road sign recognition.

Dataset contains 978 ppm format images of road sign. Train on 782 samples, validate on 196 samples.

Library Used

  • Keras
  • Sklearn
  • pandas
  • numpy

Prediction

CNN Error: 0.51%

Loss: 0.01508455599149765

Accuracy: 99.48979591836735