Man vs. computer: benchmarking machine learning algorithms for traffic sign recognition

Publikation: Bidrag til tidsskriftKonferenceartikelfagfællebedømt

Standard

Man vs. computer : benchmarking machine learning algorithms for traffic sign recognition. / Stallkamp, J.; Schlipsing, M.; Salmen, J.; Igel, Christian.

I: Neural Networks, Bind 32, 2012, s. 323-332.

Publikation: Bidrag til tidsskriftKonferenceartikelfagfællebedømt

Harvard

Stallkamp, J, Schlipsing, M, Salmen, J & Igel, C 2012, 'Man vs. computer: benchmarking machine learning algorithms for traffic sign recognition', Neural Networks, bind 32, s. 323-332. https://doi.org/10.1016/j.neunet.2012.02.016

APA

Stallkamp, J., Schlipsing, M., Salmen, J., & Igel, C. (2012). Man vs. computer: benchmarking machine learning algorithms for traffic sign recognition. Neural Networks, 32, 323-332. https://doi.org/10.1016/j.neunet.2012.02.016

Vancouver

Stallkamp J, Schlipsing M, Salmen J, Igel C. Man vs. computer: benchmarking machine learning algorithms for traffic sign recognition. Neural Networks. 2012;32:323-332. https://doi.org/10.1016/j.neunet.2012.02.016

Author

Stallkamp, J. ; Schlipsing, M. ; Salmen, J. ; Igel, Christian. / Man vs. computer : benchmarking machine learning algorithms for traffic sign recognition. I: Neural Networks. 2012 ; Bind 32. s. 323-332.

Bibtex

@inproceedings{068115b4589b4b7ba6b953ee332e7c4d,
title = "Man vs. computer: benchmarking machine learning algorithms for traffic sign recognition",
abstract = "Traffic signs are characterized by a wide variability in their visual appearance in real-world environments. For example, changes of illumination, varying weather conditions and partial occlusions impact the perception of road signs. In practice, a large number of different sign classes needs to be recognized with very high accuracy. Traffic signs have been designed to be easily readable for humans, who perform very well at this task. For computer systems, however, classifying traffic signs still seems to pose a challenging pattern recognition problem. Both image processing and machine learning algorithms are continuously refined to improve on this task. But little systematic comparison of such systems exist. What is the status quo? Do today{\textquoteright}s algorithms reach human performance? For assessing the performance of state-of-the-art machine learning algorithms, we present a publicly available traffic sign dataset with more than 50,000 images of German road signs in 43 classes. The data was considered in the second stage of the GermanTraffic Sign Recognition Benchmark held at IJCNN 2011. The results of this competition are reported and the best-performing algorithms are briefly described. Convolutional neural networks (CNNs) showed particularly high classification accuracies in the competition. We measured the performance of human subjects on the same data—and the CNNs outperformed the human test persons.",
author = "J. Stallkamp and M. Schlipsing and J. Salmen and Christian Igel",
note = "Selected Papers from IJCNN 2011; null ; Conference date: 31-07-2011 Through 05-08-2011",
year = "2012",
doi = "10.1016/j.neunet.2012.02.016",
language = "English",
volume = "32",
pages = "323--332",
journal = "Neural Networks",
issn = "0893-6080",
publisher = "Pergamon Press",

}

RIS

TY - GEN

T1 - Man vs. computer

AU - Stallkamp, J.

AU - Schlipsing, M.

AU - Salmen, J.

AU - Igel, Christian

N1 - Selected Papers from IJCNN 2011

PY - 2012

Y1 - 2012

N2 - Traffic signs are characterized by a wide variability in their visual appearance in real-world environments. For example, changes of illumination, varying weather conditions and partial occlusions impact the perception of road signs. In practice, a large number of different sign classes needs to be recognized with very high accuracy. Traffic signs have been designed to be easily readable for humans, who perform very well at this task. For computer systems, however, classifying traffic signs still seems to pose a challenging pattern recognition problem. Both image processing and machine learning algorithms are continuously refined to improve on this task. But little systematic comparison of such systems exist. What is the status quo? Do today’s algorithms reach human performance? For assessing the performance of state-of-the-art machine learning algorithms, we present a publicly available traffic sign dataset with more than 50,000 images of German road signs in 43 classes. The data was considered in the second stage of the GermanTraffic Sign Recognition Benchmark held at IJCNN 2011. The results of this competition are reported and the best-performing algorithms are briefly described. Convolutional neural networks (CNNs) showed particularly high classification accuracies in the competition. We measured the performance of human subjects on the same data—and the CNNs outperformed the human test persons.

AB - Traffic signs are characterized by a wide variability in their visual appearance in real-world environments. For example, changes of illumination, varying weather conditions and partial occlusions impact the perception of road signs. In practice, a large number of different sign classes needs to be recognized with very high accuracy. Traffic signs have been designed to be easily readable for humans, who perform very well at this task. For computer systems, however, classifying traffic signs still seems to pose a challenging pattern recognition problem. Both image processing and machine learning algorithms are continuously refined to improve on this task. But little systematic comparison of such systems exist. What is the status quo? Do today’s algorithms reach human performance? For assessing the performance of state-of-the-art machine learning algorithms, we present a publicly available traffic sign dataset with more than 50,000 images of German road signs in 43 classes. The data was considered in the second stage of the GermanTraffic Sign Recognition Benchmark held at IJCNN 2011. The results of this competition are reported and the best-performing algorithms are briefly described. Convolutional neural networks (CNNs) showed particularly high classification accuracies in the competition. We measured the performance of human subjects on the same data—and the CNNs outperformed the human test persons.

U2 - 10.1016/j.neunet.2012.02.016

DO - 10.1016/j.neunet.2012.02.016

M3 - Conference article

C2 - 22394690

VL - 32

SP - 323

EP - 332

JO - Neural Networks

JF - Neural Networks

SN - 0893-6080

Y2 - 31 July 2011 through 5 August 2011

ER -

ID: 40393629