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

By Johannes Stallkamp, Marc Schlipsing, Jan Salmen, Christian Igel. 2012

1,816 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

In Evaluation & Benchmarks

Citations in the corpus

Delving into Transferable Adversarial Examples and Black-box Attacks (2016, background): “Our work differs from Papernot et al. (2016a;b) (Stallkamp et al. (2012)); in our work, we study the transferability over larger models and a larger dataset, i.e., ImageNet.”
Deep learning in neural networks: An overview (2014, neutral): “This is illustrated by better and better results obtained by various teams over time (Stallkamp et al., 2012) (the organizers eventually imposed a limit of 10 resubmissions).”
Robust Physical-World Attacks on Deep Learning Models (2017, methodology): “We observe that a camouflage-art attack results in GTSRB*-CNN misclassifying a Stop sign as a Speed Limit 80 sign in 80% of the stationary testing conditions, and in 87.5% of the drive-by testing conditions.”
Safety Verification of Deep Neural Networks (2016, background): “[35] and 1000 classes of colour images used for the well-known imageNet large-scale visual recognition challenge (ILSVRC) [4].”
Incremental Learning Through Deep Adaptation (2017, methodology): “Third, we see a distinction between natural images (Caltech-256, CIFAR-10, SVHN, GTSR, Daimler) and unnatural ones (Sketch, Omniglot, Plankton) ; the Plankton images are natural in the rigorous sense, but do seem to exhibit different behavior than the rest.”
Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation (2017, methodology): “German Traffic Sign Benchmark: As a third dataset we used the German Traffic Sign Benchmark (GTSB) [23], which consists of images of german traffic signs, which has 43 classes with 34209 training and 12630 test samples.”
Measuring the tendency of CNNs to Learn Surface Statistical Regularities (2017, background): “With respect to related work, we high-light [17] which showed that CNNs trained on the clean MNIST, CIFAR-10 and GTSRB [39] datasets generalize quite poorly to the so-called “negative” test sets, where the test images have negated brightness intensities.”
ABS: Scanning Neural Networks for Back-doors by Artificial Brain Stimulation (2019, methodology): “In particular, the trojan size is around 6% (of the input size) in our CIFAR and GTSRB attacks and the authors reported that NC is less effective when the trigger size goes beyond 6%.”
Deep Learning for Large-Scale Traffic-Sign Detection and Recognition (2019, background): “It has been shown in [37] that the performance of CNN on recognition outperforms the human performance on GTSRB.”
Traffic sign recognition — How far are we from the solution? (2013, methodology): “The training times for the projections on the GTSC are summarized in Table IV.”