Description of Corner Cases in Automated Driving: Goals and Challenges

Bogdoll, Daniel; Breitenstein, Jasmin; Heidecker, Florian; Bieshaar, Maarten; Sick, Bernhard; Fingscheidt, Tim; Zöllner, J. Marius

doi:10.1109/iccvw54120.2021.00119

Cited by 36 publications

(15 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further, two out-of-domain corner case situations, as described in [8], were used as input for the KITTI trained models to test the stability of the approaches. The first case is a night scene from the Nighttime Driving dataset [13] and the second case is a scene from the Fishyscapes benchmark [7] that has an unconventional object positioned in a street scene.…”

Section: Image Compressionmentioning

confidence: 99%

Compressing Sensor Data for Remote Assistance of Autonomous Vehicles using Deep Generative Models

Bogdoll¹,

Johannes²,

Rauch³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

In the foreseeable future, autonomous vehicles will require human assistance in situations they can not resolve on their own. In such scenarios, remote assistance from a human can provide the required input for the vehicle to continue its operation. Typical sensors used in autonomous vehicles include camera and lidar sensors. Due to the massive volume of sensor data that must be sent in real-time, highly efficient data compression is elementary to prevent an overload of network infrastructure. Sensor data compression using deep generative neural networks has been shown to outperform traditional compression approaches for both image and lidar data, regarding compression rate as well as reconstruction quality. However, there is a lack of research about the performance of generative-neural-network-based compression algorithms for remote assistance. In order to gain insights into the feasibility of deep generative models for usage in remote assistance, we evaluate state-of-the-art algorithms regarding their applicability and identify potential weaknesses. Further, we implement an online pipeline for processing sensor data and demonstrate its performance for remote assistance using the CARLA simulator.

show abstract

Section: Image Compressionmentioning

confidence: 99%

Compressing Sensor Data for Remote Assistance of Autonomous Vehicles using Deep Generative Models

Bogdoll¹,

Johannes²,

Rauch³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…As these models rely on large training datasets, they typically assume that all classes that are to be detected have been present during training [7]. If the model encounters instances that are outside the training distribution of the network, also called anomalies or corner cases [8], they tend to fail [6]. Even by defining an Operational Design Domain (ODD), the full set of objects that can occur on roads is not predictable.…”

Section: Introductionmentioning

confidence: 99%

Multimodal Detection of Unknown Objects on Roads for Autonomous Driving

Bogdoll¹,

Enrico²,

Nitsche³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Tremendous progress in deep learning over the last years has led towards a future with autonomous vehicles on our roads. Nevertheless, the performance of their perception systems is strongly dependent on the quality of the utilized training data. As these usually only cover a fraction of all object classes an autonomous driving system will face, such systems struggle with handling the unexpected. In order to safely operate on public roads, the identification of objects from unknown classes remains a crucial task. In this paper, we propose a novel pipeline to detect unknown objects. Instead of focusing on a single sensor modality, we make use of lidar and camera data by combining state-of-the art detection models in a sequential manner. We evaluate our approach on the Waymo Open Perception Dataset and point out current research gaps in anomaly detection.

show abstract

“…While strong violations of rules are unacceptable, slight twists regarding the rule conformity (RC) are socially accepted. Especially in corner case situations [1], a flexibility that goes beyond enforced traffic rules is often necessary. Unfortunately, this knowledge about human behavior is not freely available but can be extracted from datasets.…”

Section: Introductionmentioning

confidence: 99%

Quantification of Actual Road User Behavior on the Basis of Given Traffic Rules

Bogdoll,

Nekolla,

Joseph

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Driving on roads is restricted by various traffic rules, aiming to ensure safety for all traffic participants. However, human road users usually do not adhere to these rules strictly, resulting in varying degrees of rule conformity. Such deviations from given rules are key components of today's road traffic. In autonomous driving, robotic agents can disturb traffic flow, when rule deviations are not taken into account. In this paper, we present an approach to derive the distribution of degrees of rule conformity from human driving data. We demonstrate our method with the Waymo Open Motion dataset and Safety Distance and Speed Limit rules.

show abstract

Description of Corner Cases in Automated Driving: Goals and Challenges

Cited by 36 publications

References 21 publications

Compressing Sensor Data for Remote Assistance of Autonomous Vehicles using Deep Generative Models

Compressing Sensor Data for Remote Assistance of Autonomous Vehicles using Deep Generative Models

Multimodal Detection of Unknown Objects on Roads for Autonomous Driving

Quantification of Actual Road User Behavior on the Basis of Given Traffic Rules

Contact Info

Product

Resources

About