DAROD: A Deep Automotive Radar Object Detector on Range-Doppler maps

Colin, Decourt,; VanRullen, Rufin; Salle, Didier; Oberlin, Thomas

doi:10.1109/iv51971.2022.9827281

Cited by 13 publications

(4 citation statements)

References 65 publications

(122 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our study, we choose three IoU threshold values (0.4, 0.45, and 0.5) and compare the results to determine the best value that suits our work. There values are the frequently used in the literature [58], [59]. In fact, when the IoU threshold is adjusted from 0.5 to slightly lower values, such as 0.45 and 0.40, the detector tends to predict more small objects included in the training dataset, and thus the detection performance will be improved.…”

Section: Implementation and Resultsmentioning

confidence: 99%

Classification and Localization of Arabic Handwritten Text in KHATT Dataset Based on Faster R-CNN

May Mowaffaq AL-Taee

2023

pst

View full text Add to dashboard Cite

Arabic handwriting recognition is an important research area in computer vision. Due to the complexity of the Arabic script, this is an arduous task. Several approaches have been proposed to address this challenge, including deep learning algorithms. Despite their efficiency, these approaches present some limitations such as the use of lexicon-driven models, the need of a lot of data for training and the huge computational cost. We propose, in this paper, two novel models based on the robust Faster Region-Convolution Neural Network (Faster R-CNN) to recognize Arabic handwritten sentences. The Faster R-CNN is commonly used to detect objects in images and depends on region proposal algorithms. These models use the pre-trained VGG 16 and ResNet50. Moreover, they use a soft non-maximum suppression strategy (Soft-NMS) in the post-processing stage instead of the traditional NMS to increase the detection rate of overlapping items. The models will be trained independently to identify words in the text lines and tested using sentences from the (KFUPM Handwritten Arabic TexT) KHATT dataset. It will be shown that the faster R-CNN models allow greater accuracy and effectiveness in handwriting recognition. They achieve accuracy rates of 99% and 98% for the two networks, VGG16 and ResNet50, respectively

show abstract

Section: Implementation and Resultsmentioning

confidence: 99%

Classification and Localization of Arabic Handwritten Text in KHATT Dataset Based on Faster R-CNN

May Mowaffaq AL-Taee

2023

pst

View full text Add to dashboard Cite

show abstract

“…Their approach is further improved both in performance and computation time by Ju et al in [14]. Decourt et al [15] work on range-Doppler spectra to solve an object detection task. Rebut et al [16] also use range-Doppler spectra, but in their approach the angular representation is learned as part of the object detection network.…”

Section: Related Workmentioning

confidence: 99%

Histogram-based Deep Learning for Automotive Radar

Tatarchenko¹,

Rambach²

2023

Preprint

View full text Add to dashboard Cite

There are various automotive applications that rely on correctly interpreting point cloud data recorded with radar sensors. We present a deep learning approach for histogrambased processing of such point clouds. Compared to existing methods, the design of our approach is extremely simple: it boils down to computing a point cloud histogram and passing it through a multi-layer perceptron. Our approach matches and surpasses state-of-the-art approaches on the task of automotive radar object type classification. It is also robust to noise that often corrupts radar measurements, and can deal with missing features of single radar reflections. Finally, the design of our approach makes it more interpretable than existing methods, allowing insightful analysis of its decisions.

show abstract

“…Radar data can be represented as a list of targets (point cloud) or as a raw data tensor (range-Doppler or range-angle-Doppler map). The target list is the default radar data format and contains very low-level information such as the position, velocity, and radar cross-section [4] of targets around the vehicle. Point cloud data are usually processed by LiDAR, and information will be lost in harsh environments; the original data tensor is a relatively low-level and has a widely used data format, and will not lose a lot of information, so in this paper, radar range-Doppler (RD) maps are used as input for object detection.…”

Section: Introductionmentioning

confidence: 99%

Radar Timing Range–Doppler Spectral Target Detection Based on Attention ConvLSTM in Traffic Scenes

Jia,

Tan,

et al. 2023

Remote Sensing

View full text Add to dashboard Cite

With the development of autonomous driving and the emergence of various intelligent traffic scenarios, object detection technology based on deep learning is more and more widely applied to real traffic scenarios. Commonly used detection devices include LiDAR and cameras. Since the implementation of traffic scene target detection technology requires mass production, the advantages of millimeter-wave radar have emerged, such as low cost and no interference from the external environment. The performance of LiDAR and cameras is greatly reduced due to their sensitivity to light, which affects target detection at night and in bad weather. However, millimeter-wave radar can overcome the influence of these harsh environments and has a great auxiliary effect on safe driving on the road. In this work, we propose a deep-learning-based object detection method considering the radar range–Doppler spectrum in traffic scenarios. The algorithm uses YOLOv8 as the basic architecture, makes full use of the time series characteristics of range–Doppler spectrum data in traffic scenarios, introduces the ConvLSTM network, and exerts the ability to process time series data. In order to improve the model’s ability to detect small objects, an efficient and lightweight Efficient Channel Attention (ECA) module is introduced. Through extensive experiments, our model shows better performance on two publicly available radar datasets, CARRADA and RADDet, compared to other state-of-the-art methods. Compared with other mainstream methods that can only achieve 30–60% mAP performance when the IOU is 0.3, our model can achieve 74.51% and 75.62% on the RADDet and CARRADA datasets, respectively, and has better robustness and generalization ability.

show abstract

DAROD: A Deep Automotive Radar Object Detector on Range-Doppler maps

Cited by 13 publications

References 65 publications

Classification and Localization of Arabic Handwritten Text in KHATT Dataset Based on Faster R-CNN

Classification and Localization of Arabic Handwritten Text in KHATT Dataset Based on Faster R-CNN

Histogram-based Deep Learning for Automotive Radar

Radar Timing Range–Doppler Spectral Target Detection Based on Attention ConvLSTM in Traffic Scenes

Contact Info

Product

Resources

About