Fig 5 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
Despite the high quality performance of the deep neural network in real-world applications, they are susceptible to minor perturbations of adversarial attacks. This is mostly undetectable to human vision. The impact of such attacks has become extremely detrimental in autonomous vehicles with real-time "safety" concerns. The black-box adversarial at...
Context in source publication
Context 1
... (GTSRB) dataset [24]. It consists of 43 traffic sign classes, where 39000 are training images and 12000 are test images. The images contain one traffic sign, a border of 10% around the actual traffic sign (at least 5 pixels) to allow for edge-based approaches. It varies between (15 × 15) to (250 × 250) pixels and sample images are shown in Fig. ...
Similar publications
Semantic segmentation enables vehicles to accurately identify and categorize objects in their surroundings, such as pedestrians, other cars, road signs, and obstacles. Semantic segmentation, object detection, and deep learning have emerged as critical pillars, enabling machines to perceive and understand the visual world with unprecedented precisio...
Autonomous vehicles rely on LiDAR sensors to detect obstacles such as pedestrians, other vehicles, and fixed infrastructures. LiDAR spoofing attacks have been demonstrated that either create erroneous obstacles or prevent detection of real obstacles, resulting in unsafe driving behaviors. In this paper, we propose an approach to detect and mitigate...
![Fig. 4: Comparison of IoUMOS on the SemanticKITTI [3] validation set...](profile/Benedikt-Mersch-2/publication/361338263/figure/fig1/AS:1169324840628227@1655800175105/Comparison-of-IoUMOS-on-the-SemanticKITTI-3-validation-set-using-different-moving_Q320.jpg)
A key challenge for autonomous vehicles is to navigate in unseen dynamic environments. Separating moving objects from static ones is essential for navigation, pose estimation, and understanding how other traffic participants are likely to move in the near future. In this work, we tackle the problem of distinguishing 3D LiDAR points that belong to c...
![Fig. 4: Comparison of IoUMOS on the SemanticKITTI [3] validation set...](publication/361206305/figure/fig1/AS:1165261960953856@1654831509202/Comparison-of-IoUMOS-on-the-SemanticKITTI-3-validation-set-using-different-moving_Q320.jpg)
A key challenge for autonomous vehicles is to navigate in unseen dynamic environments. Separating moving objects from static ones is essential for navigation, pose estimation, and understanding how other traffic participants are likely to move in the near future. In this work, we tackle the problem of distinguishing 3D LiDAR points that belong to c...
Despite autonomous vehicles (AVs) being safer than human drivers, people are averse to their presence on roads. Across three studies (N = 4,014), we examined peoples’ perceptions of human drivers and AVs acting within a moral dilemma. Scenarios involved an out-of-control vehicle (piloted by a human, or autonomously) that could stay on its present c...
