Probably Unknown: Deep Inverse Sensor Modelling In Radar

Weston, Rob; Cen, Sarah H.; Newman, P. R.; Posner, Ingmar

doi:10.48550/arxiv.1810.08151

Cited by 5 publications

(14 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Occupancy grid methods [11] provide discrimination between free and occupied space, dealing with the complex sensor artefacts encountered in radar scans, but fail to deliver richer information regarding the nature of the occupied space. To overcome this, a FCNN is used in [12] to segment a radar scan based on the probability of occupancy and hand labels are used to classify objects.…”

Section: Related Workmentioning

confidence: 99%

“…An example of both inexact and inaccurate supervision applied to radar is [11], where the authors use a U-Net, the popular image segmentation network [17], to segment a radar scan based on probability of occupancy on a more extensive dataset and produces good results. In contrast to [11], however, where the shorter maximum range of the LiDAR is not dealt with, we accumulate LiDAR returns in the native radar representation using a good external source of egomotion.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

RSS-Net: Weakly-Supervised Multi-Class Semantic Segmentation with FMCW Radar

Kaul¹,

Martini²,

Gadd³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

This paper presents an efficient annotation procedure and an application thereof to end-to-end, rich semantic segmentation of the sensed environment using Frequency-Modulated Continuous-Wave scanning radar. We advocate radar over the traditional sensors used for this task as it operates at longer ranges and is substantially more robust to adverse weather and illumination conditions. We avoid laborious manual labelling by exploiting the largest radar-focused urban autonomy dataset collected to date, correlating radar scans with RGB cameras and LiDAR sensors, for which semantic segmentation is an already consolidated procedure. The training procedure leverages a stateof-the-art natural image segmentation system which is publicly available and as such, in contrast to previous approaches, allows for the production of copious labels for the radar stream by incorporating four camera and two LiDAR streams. Additionally, the losses are computed taking into account labels to the radar sensor horizon by accumulating LiDAR returns along a posechain ahead and behind of the current vehicle position. Finally, we present the network with multi-channel radar scan inputs in order to deal with ephemeral and dynamic scene objects.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

RSS-Net: Weakly-Supervised Multi-Class Semantic Segmentation with FMCW Radar

Kaul¹,

Martini²,

Gadd³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…It should also encourage sensor fusion in temporal data. Secondly, we describe a semiautomatic method to generate the radar annotations by only using the camera information instead of a lidar as usual [25], [27], [38]. The aim of this contribution is to reduce annotation time and cost by exploiting visual information without the need for an expensive sensor.…”

Section: Introductionmentioning

confidence: 99%

CARRADA Dataset: Camera and Automotive Radar with Range-Angle-Doppler Annotations

Ouaknine

Newson

Rebut

et al. 2020

Preprint

View full text Add to dashboard Cite

High quality perception is essential for autonomous driving (AD) systems. To reach the accuracy and robustness that are required by such systems, several types of sensors must be combined. Currently, mostly cameras and laser scanners (lidar) are deployed to build a representation of the world around the vehicle. While radar sensors have been used for a long time in the automotive industry, they are still under-used for AD despite their appealing characteristics (notably, their ability to measure the relative speed of obstacles and to operate even in adverse weather conditions). To a large extent, this situation is due to the relative lack of automotive datasets with real radar signals that are both raw and annotated. In this work, we introduce CARRADA, a dataset of synchronized camera and radar recordings with rangeangle-Doppler annotations. We also present a semi-automatic annotation approach, which was used to annotate the dataset, and a radar semantic segmentation baseline, which we evaluate on several metrics. Both our code and dataset will be released.

show abstract

“…Hence, the information provided by two separate measurements is largely independent. In contrast to that, deep ISMs do also infer the occupancy state of areas 1 Daniel Bauer and Lars Kuhnert are with Ford Werke GmbH, Germany {dbauer31, lkuhnert}@ford.com 2 Lutz Eckstein is with the Institute of Automotive Engineering, RWTH Aachen University, Germany office@ika.rwth-aachen.de far away from actual detections. This leads to situations in which two independent measurements can provide mostly the same information about a cell's state.…”

Section: Introductionmentioning

confidence: 99%

“…multipath reflections). Recently, deep learning-based inverse sensor models, from here on called deep ISMs, have been shown to improve over their geometric counterparts in retrieving occupancy information [1], [2], [3]. Nevertheless, these methods perform a data-driven interpolation which has to be verified later on in the presence of measurements.…”

mentioning

confidence: 99%