Archive Meta Data

Author(s)

Johannes Stallkamp, Marc Schlipsing, Jan Salmen, Christian Igel

Title

German Traffic Sign Recognition Benchmark GTSRB

Description

The German Traffic Sign Recognition Benchmark (GTSRB) is a multi-class image classification benchmark in the domain of advanced driver assistance systems and autonomous driving. 
It was first published at IJCNN 2011. 

The official training data (use this to train your model):
 - Images and annotations (GTSRB_Final_Training_Images.zip)
 - Three sets of different HOG features (GTSRB_Final_Training_HOG.zip)
 - Haar-like features (GTSRB_Final_Training_Haar.zip)
 - Hue histograms (GTSRB_Final_Training_HueHist.zip)

The official test dataset (use this to test your model):
 - Images and annotations (without ground truth classes) (GTSRB_Final_Test_Images.zip)
 - Extended ground truth annotations (with classes) (GTSRB_Final_Test_GT.zip)
 - Three sets of different HOG features (GTSRB_Final_Test_HOG.zip)
 - Haar-like features (GTSRB_Final_Test_Haar.zip)
 - Hue histograms (GTSRB_Final_Test_HueHist.zip)
 
Training dataset that was used at the IJCNN 2011 competition (for reproducibility):
 - Images and annotations (GTSRB-Training_fixed.zip)
 - Three sets of different HOG features (GTSRB_Training_Features_HOG.zip)
 - Haar-like features (GTSRB_Training_Features_Haar.zip)
 - Hue histograms (GTSRB_Training_Features_HueHist.zip)

Test dataset that was used at the IJCNN 2011 competition (for reproducibility):
 - Images and annotations (GTSRB_Online-Test-Images.zip)
 - Three sets of different HOG features (GTSRB_Online-Test-HOG.zip)
 - Haar-like features (GTSRB_Online-Test-Haar.zip)
 - Hue histograms (GTSRB_Online-Test-HueHist.zip)

Test dataset that was used at the IJCNN 2011 competition (for reproducibility, elements are sorted by class and track):
 - Images and annotations (GTSRB_Online-Test-Images-Sorted.zip)
 - Three sets of different HOG features (GTSRB_Online-Test-HOG-Sorted.zip)
 - Haar-like features (GTSRB_Online-Test-Haar-Sorted.zip)
 - Hue histograms (GTSRB_Online-Test-HueHist-Sorted.zip)


References:
Johannes Stallkamp, Marc Schlipsing, Jan Salmen, and Christian Igel. Man vs. Computer: Benchmarking Machine Learning Algorithms for Traffic Sign Recognition. Neural Networks32, pp. 323-332, 2012
Johannes Stallkamp, Marc Schlipsing, Jan Salmen, and Christian Igel. The German Traffic Sign Recognition Benchmark: A Multi-class Classification Competition. International Joint Conference on Neural Networks (IJCNN 2011), pp. 1453-1460, IEEE Press