Evaluation of Traffic Sign Recognition Methods Trained on Synthetically Generated Data

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8192)


Most of today’s machine learning techniques requires large manually labeled data. This problem can be solved by using synthetic images. Our main contribution is to evaluate methods of traffic sign recognition trained on synthetically generated data and show that results are comparable with results of classifiers trained on real dataset. To get a representative synthetic dataset we model different sign image variations such as intra-class variability, imprecise localization, blur, lighting, and viewpoint changes. We also present a new method for traffic sign segmentation, based on a nearest neighbor search in the large set of synthetically generated samples, which improves current traffic sign recognition algorithms.


synthetic data traffic sign recognition nearest neighbor search 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Stallkamp, J., Schlipsing, M., Salmen, J., Igel, C.: Man vs. Computer: Benchmarking Machine Learning Algorithms for Traffic Sign Recognition. Neural Networks (2012)Google Scholar
  2. 2.
    Paclik, P., Novovicova, J., Duin, R.: Building Road-SignClassifiers Using a Trainable Similarity Measure. IEEE Trans. Intell. Transp. Syst. 7(3), 309–321 (2006)CrossRefGoogle Scholar
  3. 3.
    Zaklouta, F., Stanciulescu, B., Hamdoun, O.: Traffic sign classification using K-d trees and Random Forests. In: IEEE International Joint Conference on Neural Networks, San Jose, California, pp. 2151–2155 (2011)Google Scholar
  4. 4.
    Maldonado-Bascón, S., Lafuente-Arroyo, S., Gil-Jimenez, P., Gomez-Moreno, H., Lopez-Ferreras, F.: Road-Sign Detection and Recognition Based on Support Vector Machines. IEEE Trans. Intell. Transp. Syst. 8(2), 264–278 (2007)CrossRefzbMATHGoogle Scholar
  5. 5.
    Timofte, R., Zimmermann, K., Gool, L.V.: Multi-view traffic sign detection, recognition, and 3D localization. In: Workshop on Applications of Computer Vision, Snowbird, Utah, pp. 1–8 (2009)Google Scholar
  6. 6.
    Ciresan, D., Meier, U., Masci, J., Schmindhuber, J.: A Committee of Neural Networks for Traffic Sign Classification. In: IEEE International Joint Conference on Neural Networks, San Jose, California, pp. 1918–1921 (2011)Google Scholar
  7. 7.
    Sermanet, P., Lecun, Y.: Traffic Sign Recognition with Multi-Scale Convolutional Networks. In: International Joint Conference on Neural Networks, San Jose, California, pp. 2809–2813 (2011)Google Scholar
  8. 8.
    Shotton, J., Fitzgibbon, A., Cook, M., Sharp, T., Finocchio, M., Moore, R., Kipman, A., Blake, A.: Real-Time Human Pose Recognition in Parts from a Single Depth Image. In: Proceedings IEEE Computer Vision and Pattern Recognition, Colorado, USA, pp. 1297–1304 (2011)Google Scholar
  9. 9.
    Grauman, K., Shakhnarovichand, G., Darrell, T.: Inferring 3D structure with a statistical image-based shape model. In: Proceedings Ninth IEEE International Conference on Computer Vision, Nice, France, pp. 641–647 (2003)Google Scholar
  10. 10.
    Stark, M., Goesele, M., Schiele, B.: Back to the Future: Learning Shape Models from 3D CAD Data. In: Proceedingsof the British Machine Vision Conference, Aberystwyth, Wales, pp. 106.1–106.11 (2010)Google Scholar
  11. 11.
    Marin, J., Vazquez, D., Geronimo, D., Lopez, A.M.: Learning Appearance in Virtual Scenarios for Pedestrian Detection. In: Proceedings Computer Vision and Pattern Recognition, San Francisco, California, pp. 137–144 (2010)Google Scholar
  12. 12.
    Pishchulin, L., Thorm, T., Wojek, C., Andriluka, M., Thormahlen, T., Schiele, B.: Learning People Detection Models from Few Training Samples. In: Proceedings Computer Vision and Pattern Recognition, Colorado Springs, pp. 1–8 (2011)Google Scholar
  13. 13.
    Enzweiler, M., Gavrila, D.M.: A Mixed Generative-Discriminative Framework for Pedestrian Classification. In: Proceedings Computer Vision and Pattern Recognition, Anchorage, Alaska, USA (2008)Google Scholar
  14. 14.
    Liebelt, J., Schmid, C., Schertler, K.: Viewpoint-Independent Object Class Detection using 3D Feature Maps. In: Proceedings Computer Vision and Pattern Recognition, Anchorage, Alaska, USA (2008)Google Scholar
  15. 15.
    Wang, K., Babenko, B., Belongie, S.: End-to-End Scene Text Recognition. In: International Conference on Computer Vision, Barcelona, Spain (2011)Google Scholar
  16. 16.
    Ozuysal, M., Fua, P., Lepetit, V.: Fast Keypoint Recognition in Ten Lines of Code. In: IEEE Conference on Computer Vision and Pattern Recognition, Minneapolis, Minnesota, pp. 1–8 (2007)Google Scholar
  17. 17.
    Mogelmose, A., Trivedi, M., Mouslund, T.: Learning to Detect Traffic Signs: Comparative Evaluation of Synthetic and Real-world Datasets. In: 21st International Conference on Pattern Recognition, Tsukuba, Japan (2012)Google Scholar
  18. 18.
    Overett, G.M., Tychsen-Smith, L., Petersson, L., Andersson, L., Pettersson, N.: Creating Robust High-Throughput Traffic Sign Detectors Using Centre-Surround HOG Statistics. Machine Vision and Applications Special Issue Paper, 1–14 (December 2011)Google Scholar
  19. 19.
    Medici, P., Caraffi, C., Cardarelli, E., Porta, P.P., Ghisto, G.: Real Time Road Signs Classification. In: IEEE Conference on Vehicular Electronics and Safety, Columbus, OH, USA (2008)Google Scholar
  20. 20.
    Larsson, F.: Sweden traffic signs dataset,
  21. 21.
    Larsson, F., Felsberg, M.: Using Fourier Descriptors and Spatial Models for Traffic Sign Recognition. In: Proceedings of Scandinavian Conference on Image Analysis, Ystad, Sweden, pp. 238–249 (2011)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  1. 1.Graphics and Media LabLomonosov Moscow State UniversityRussia

Personalised recommendations