High Resolution Radar Dataset for Semi-Supervised Learning of Dynamic Objects

Mostajabi, Mohammadreza; Wang, Ching Ming; Ranjan, Darsh; Hsyu, G.

doi:10.1109/cvprw50498.2020.00058

Cited by 39 publications

(17 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [14] the authors claim to provide the first high resolution radar dataset of complex urban driving scenes. The annotation was performed by unsupervised dynamic object detection, aided by a Synthetic Aperture Radar (SAR) image of the static environment created using high quality inertial sensors on the vehicle.…”

Section: Related Workmentioning

confidence: 99%

Automatic labeling of vulnerable road users in multi-sensor data

Dimitrievski

Shopovska

Hamme

et al. 2021

2021 IEEE International Intelligent Transportation Systems Conference (ITSC)

View full text Add to dashboard Cite

A growing interest in technologies for autonomous driving emphasizes the demand for safe and reliable perception systems in various driving conditions. The current state-of-theart perception solutions rely on data-driven machine learning approaches, and require large amounts of annotated data to train accurate models. In this study we have identified limitations in the existing radar-based traffic datasets, and propose a richer, annotated raw radar dataset. The proposed solution is a semi-automatic data labeling tool, which generates an initial set of candidate annotations using state-of-the-art automatic object recognition algorithms, and requires only minimal manual intervention. In the first qualitative evaluation ever for automotive radar datasets we measure the quality of automatically computed labels under various light conditions, occlusion, behavior and modeling bias based on a multitude of tracking metrics. We determined the specific cases where automatic labeling is sufficient and where a human annotator needs to inspect and manually correct errors made by the algorithms.

show abstract

Section: Related Workmentioning

confidence: 99%

Automatic labeling of vulnerable road users in multi-sensor data

Dimitrievski

Shopovska

Hamme

et al. 2021

2021 IEEE International Intelligent Transportation Systems Conference (ITSC)

View full text Add to dashboard Cite

show abstract

“…However, most of them are very small, use a special radar sensor, or have other limitations. Currently, the following open radar data sets are available: nuScenes [16], Astyx [17], Oxford Radar RobotCar [18], MulRan [19], RADIATE [20], CARRADA [21], Zendar [22], NLOS-Radar [23], and the CRUW data set [24]. Despite the increasing number of candidates, most of them are not applicable to the given problem for various reasons.…”

Section: Related Workmentioning

confidence: 99%

“…The Zendar [22] and nuScenes [16] data sets use real world recordings for a reasonable amount of different scenarios. However, their radar sensors offer an extremely sparse output for other road users which seems not reasonable for purely radar-based evaluations.…”

Section: Related Workmentioning

confidence: 99%

RadarScenes: A Real-World Radar Point Cloud Data Set for Automotive Applications

Schumann¹,

Hahn²,

Scheiner³

et al. 2021

Preprint

View full text Add to dashboard Cite

“…Automotive radar, also known as Frequency Modulated Continuous Wave (FMCW) radar, has been widely adopted in Advanced Driver Assistance Systems (ADAS). Compared to other sensors such as camera and Lidar, radar is robust against adverse weather and lighting conditions, and is a less expensive [1] sensor. The drawbacks of radar sensor are its low output resolution and high noise levels.…”

Section: Introductionmentioning

confidence: 99%

“…• A novel dataset that contains radar data in the form of RAD representations with the corresponding annotations for various object classes is introduced. The dataset is available 1 for future research in this field.…”

Section: Introductionmentioning

confidence: 99%

RADDet: Range-Azimuth-Doppler based Radar Object Detection for Dynamic Road Users

Zhang

Nowruzi

Laganière

2021

Preprint

View full text Add to dashboard Cite

Object detection using automotive radars has not been explored with deep learning models in comparison to the camera based approaches. This can be attributed to the lack of public radar datasets. In this paper, we collect a novel radar dataset that contains radar data in the form of Range-Azimuth-Doppler tensors along with the bounding boxes on the tensor for dynamic road users, category labels, and 2D bounding boxes on the Cartesian Bird-Eye-View range map. To build the dataset, we propose an instance-wise auto-annotation method. Furthermore, a novel Range-Azimuth-Doppler based multiclass object detection deep learning model is proposed. The algorithm is a one-stage anchor-based detector that generates both 3D bounding boxes and 2D bounding boxes on Range-Azimuth-Doppler and Cartesian domains, respectively. Our proposed algorithm achieves 56.3% AP with IOU of 0.3 on 3D bounding box predictions, and 51.6% with IOU of 0.5 on 2D bounding box prediction. Our dataset and the code can be found at https://github.com/ZhangAoCanada/RADDet.git.

show abstract

High Resolution Radar Dataset for Semi-Supervised Learning of Dynamic Objects

Cited by 39 publications

References 6 publications

Automatic labeling of vulnerable road users in multi-sensor data

Automatic labeling of vulnerable road users in multi-sensor data

RadarScenes: A Real-World Radar Point Cloud Data Set for Automotive Applications

RADDet: Range-Azimuth-Doppler based Radar Object Detection for Dynamic Road Users

Contact Info

Product

Resources

About