Traffic sign detection for U.S. roads: Remaining challenges and a case for tracking

58 citations recorded by the source corpus.

How researchers see this paper

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

SentiNet: Detecting Localized Universal Attacks Against Deep Learning Systems (2018, methodology): “a) Datasets: For this evaluation, the benign images dataset contains 145 images of stop signs from the LISA dataset [38] by cropping out the images with their labeled bounding boxes.”
Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks (2018, background): “traffic signs dataset [32] containing 6889 training and 1724 test images with bounding boxes around traffic signs and corresponding ground-truth labels.”
BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain (2017, methodology): “Thebaseline F-RCNN network is trained on the U.S. trafﬁc signs dataset [40] containing 8612 images, along with bounding boxes and ground-truth labels for each image.”
ABS: Scanning Neural Networks for Back-doors by Artificial Brain Stimulation (2019, methodology): “USTS [42] is another traffic sign dataset but used for object detection.”
BadNets: Evaluating Backdooring Attacks on Deep Neural Networks (2019, background): “traffic signs dataset [52] containing 8612 images, along with bounding boxes and ground-truth labels for each image.”
SentiNet: Detecting Physical Attacks Against Deep Learning Systems (2018, unknown): “dataset [33] by cropping out the images with their labelled bounding boxes, and place the backdoor at approximately below the “stop” text of each sign to create an “attack” dataset.”