Traffic sign recognition with multi-scale Convolutional Networks

We apply Convolutional Networks (ConvNets) to the task of traffic sign classification as part of the GTSRB competition. ConvNets are biologically-inspired multi-stage architectures that automatically learn hierarchies of invariant features. While many popular vision approaches use hand-crafted features such as HOG or SIFT, ConvNets learn features at every level from data that are tuned to the task at hand. The traditional ConvNet architecture was modified by feeding 1st stage features in addition to 2nd stage features to the classifier. The system yielded the 2nd-best accuracy of 98.97% during phase I of the competition (the best entry obtained 98.98%), above the human performance of 98.81%, using 32×32 color input images. Experiments conducted after phase 1 produced a new record of 99.17% by increasing the network capacity, and by using greyscale images instead of color. Interestingly, random features still yielded competitive results (97.33%).

[1]  Luis Moreno,et al.  Road traffic sign detection and classification , 1997, IEEE Trans. Ind. Electron..

[2]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[3]  Jana Novovičová,et al.  Road Sign Classification without Color Information , 2000 .

[4]  J. Torresen,et al.  Efficient recognition of speed limit signs , 2004, Proceedings. The 7th International IEEE Conference on Intelligent Transportation Systems (IEEE Cat. No.04TH8749).

[5]  S. Lafuente-Arroyo,et al.  Traffic sign shape classification evaluation I: SVM using distance to borders , 2005, IEEE Proceedings. Intelligent Vehicles Symposium, 2005..

[6]  Yok-Yen Nguwi,et al.  Detection and classification of road signs in natural environments , 2008, Neural Computing and Applications.

[7]  C. Bahlmann,et al.  Real-time recognition of U.S. speed signs , 2008, 2008 IEEE Intelligent Vehicles Symposium.

[8]  Eero P. Simoncelli,et al.  Nonlinear image representation using divisive normalization , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[9]  Nicolas Pinto,et al.  Why is Real-World Visual Object Recognition Hard? , 2008, PLoS Comput. Biol..

[10]  Luke Fletcher,et al.  Real-Time Speed Sign Detection Using the Radial Symmetry Detector , 2008, IEEE Transactions on Intelligent Transportation Systems.

[11]  Yann LeCun,et al.  What is the best multi-stage architecture for object recognition? , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[12]  Yann LeCun,et al.  EBLearn: Open-Source Energy-Based Learning in C++ , 2009, 2009 21st IEEE International Conference on Tools with Artificial Intelligence.

[13]  Yihong Gong,et al.  Human Tracking Using Convolutional Neural Networks , 2010, IEEE Transactions on Neural Networks.

[14]  Berin Martini,et al.  Hardware accelerated convolutional neural networks for synthetic vision systems , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[15]  Johannes Stallkamp,et al.  The German Traffic Sign Recognition Benchmark: A multi-class classification competition , 2011, The 2011 International Joint Conference on Neural Networks.