Filters








4,091 Hits in 4.2 sec

Multi-View Radar Semantic Segmentation [article]

Arthur Ouaknine, Alasdair Newson, Patrick Pérez, Florence Tupin, Julien Rebut
2021 arXiv   pre-print
In this work, we propose several novel architectures, and their associated losses, which analyse multiple "views" of the range-angle-Doppler radar tensor to segment it semantically.  ...  Fortunately, recent open-sourced datasets have opened up research on classification, object detection and semantic segmentation with raw radar signals using end-to-end trainable models.  ...  Acknowledgements We thank Veronica Elizabeth Vargas Salas for her valuable help with temporal radar data.  ... 
arXiv:2103.16214v2 fatcat:7dsyn6nfijflnp7hapzgeoilmq

Deep Multi-modal Object Detection and Semantic Segmentation for Autonomous Driving: Datasets, Methods, and Challenges [article]

Di Feng, Christian Haase-Schuetz, Lars Rosenbaum, Heinz Hertlein, Claudius Glaeser, Fabian Timm, Werner Wiesbeck, Klaus Dietmayer
2020 arXiv   pre-print
This review paper attempts to systematically summarize methodologies and discuss challenges for deep multi-modal object detection and semantic segmentation in autonomous driving.  ...  To this end, we first provide an overview of on-board sensors on test vehicles, open datasets, and background information for object detection and semantic segmentation in autonomous driving research.  ...  Deep Multi-modal Semantic Segmentation Compared to the object detection problem summarized in Sec.  ... 
arXiv:1902.07830v4 fatcat:or6enjxktnamdmh2yekejjr4re

Deep Instance Segmentation with Automotive Radar Detection Points [article]

Jianan Liu, Weiyi Xiong, Liping Bai, Yuxuan Xia, Tao Huang, Wanli Ouyang, Bing Zhu
2022 arXiv   pre-print
In this paper, we propose an efficient method based on clustering of estimated semantic information to achieve instance segmentation for the sparse radar detection points.  ...  With the development of automotive radar technologies in recent years, instance segmentation becomes possible by using automotive radar.  ...  Each radar covers a ±60 • field of view. The illustration of mounting positions of four radars and the corresponding field of view (FOV) can be seen in Fig. 5 .  ... 
arXiv:2110.01775v6 fatcat:ryhi73z53bhdldmcglq7ucgz5m

Deep learning for radar data exploitation of autonomous vehicle [article]

Arthur Ouaknine
2022 arXiv   pre-print
This thesis then present a proposed set of deep learning architectures with their associated loss functions for RADAR semantic segmentation.  ...  A deep learning architecture is also proposed to estimate the RADAR signal processing pipeline while performing multitask learning for object detection and free driving space segmentation.  ...  Multi-view RADAR semantic segmentation Motivations In this section, an approach to multi-view RADAR semantic segmentation is proposed, illustrated in Figure 5 .1, that exploits the entire data while  ... 
arXiv:2203.08038v1 fatcat:zjupxkpaffgavm45oqpwnhkczq

FISHING Net: Future Inference of Semantic Heatmaps In Grids [article]

Noureldin Hendy, Cooper Sloan, Feng Tian, Pengfei Duan, Nick Charchut, Yuesong Xie, Chuang Wang, James Philbin
2020 arXiv   pre-print
In this work, we present an end-to-end pipeline that performs semantic segmentation and short term prediction using a top-down representation.  ...  Modern autonomous robots employ multiple sets of sensors, including lidars, radars, and cameras.  ...  Related Work Semantic Segmentation In semantic segmentation a neural network is trained to predict a mask which assigns a semantic label to each pixel in the input image.  ... 
arXiv:2006.09917v1 fatcat:nlpsa74odvgddlkielutn3zimu

Raw High-Definition Radar for Multi-Task Learning [article]

Julien Rebut, Arthur Ouaknine, Waqas Malik, Patrick Pérez
2022 arXiv   pre-print
FFT-RadNet is trained both to detect vehicles and to segment free driving space. On both tasks, it competes with the most recent radar-based models while requiring less compute and memory.  ...  This unique dataset, nick-named RADIal for "Radar, Lidar et al.", is available at https://github.com/valeoai/RADIal.  ...  Moreover, there is no previous work either on free driving space segmentation or semantic segmentation using only RD views of HD radar signals.  ... 
arXiv:2112.10646v3 fatcat:pt32p4jwbzhlxdsf7bbjo6hute

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

Prannay Kaul, Daniele De Martini, Matthew Gadd, Paul Newman
2020 arXiv   pre-print
This paper presents an efficient annotation procedure and an application thereof to end-to-end, rich semantic segmentation of the sensed environment using FMCW scanning radar.  ...  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  ...  Figure 4 . 4 Overview of the network used for radar semantic segmentation. L-number of classes.  ... 
arXiv:2004.03451v1 fatcat:oonp3oek6nhvxlgzgegb3tsumm

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

Prannay Kaul, Daniele de Martini, Matthew Gadd, Paul Newman
2020 2020 IEEE Intelligent Vehicles Symposium (IV)  
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  ...  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  ...  Figure 4 . 4 Overview of the network used for radar semantic segmentation. L-number of classes.  ... 
doi:10.1109/iv47402.2020.9304674 fatcat:fj34rl7tezb6nb3lwepdwzvaam

2D Car Detection in Radar Data with PointNets [article]

Andreas Danzer, Thomas Griebel, Martin Bach, Klaus Dietmayer
2019 arXiv   pre-print
To this end, PointNets are adjusted for radar data performing 2D object classification with segmentation, and 2D bounding box regression in order to estimate an amodal 2D bounding box.  ...  Modern high-resolution radar sensors generate multiple radar targets per object, which makes these sensors particularly suitable for the 2D object detection task.  ...  [14] use the same PointNet++ architecture for semantic segmentation on radar point clouds.  ... 
arXiv:1904.08414v2 fatcat:weji5tmchrdobewzdwav6t4taq

Neural network-based semantic segmentation model for robot perception of driverless vision

Lu Ye, Ting Duan, Jiayi Zhu
2020 IET Cyber-Systems and Robotics  
To improve the recognition accuracy of the computer and enhance the ability of segmentation, in this study, depth estimation is used to predict depth information to assist semantic segmentation, and then  ...  A neural network-based semantic segmentation model is proposed. Finally, the intrinsic mechanism of attention is used to increase the correlation between channels.  ...  Taking threedimensional (3D) street view reconstruction segmentation, CSPC data set [11] is to add 3D point cloud information to the direction of semantic segmentation to obtain more accurate driverless  ... 
doi:10.1049/iet-csr.2020.0040 fatcat:cqj3ugznhzhwrp3pofea55uh7m

V2X-Sim: A Virtual Collaborative Perception Dataset for Autonomous Driving [article]

Yiming Li, Ziyan An, Zixun Wang, Yiqi Zhong, Siheng Chen, Chen Feng
2022 arXiv   pre-print
multi-modality perception, 3) diverse well-annotated ground truth to support various downstream tasks including detection, tracking, and segmentation.  ...  We seek to inspire research on multi-agent multi-modality multi-task perception, and our virtual dataset is promising to promote the development of collaborative perception before realistic datasets become  ...  Figure 2 : 2 Figure 2: Example of multi-agent multi-modality perception. From top to bottom: RGB image, depth, semantic segmentation, and BEV semantic segmentation.  ... 
arXiv:2202.08449v1 fatcat:zkhc4ibpiralhkxdj22pqpl6u4

Contrastive Learning for Automotive mmWave Radar Detection Points Based Instance Segmentation [article]

Weiyi Xiong, Jianan Liu, Yuxuan Xia, Tao Huang, Bing Zhu, Wei Xiang
2022 arXiv   pre-print
To address this issue, we propose a contrastive learning approach for implementing radar detection points-based instance segmentation.  ...  Deep learning-based instance segmentation enables real-time object identification from the radar detection points. In the conventional training process, accurate annotation is the key.  ...  segmentation with radar detection points, thus improving the performance of radar detection points based instance segmentation by applying the semantic segmentation based clustering, under the constraint  ... 
arXiv:2203.06553v1 fatcat:vifbvx2jxndn3c4pumjyy2fqla

Multi-Modal 3D Object Detection in Autonomous Driving: a Survey [article]

Yingjie Wang, Qiuyu Mao, Hanqi Zhu, Yu Zhang, Jianmin Ji, Yanyong Zhang
2021 arXiv   pre-print
So far, there has been no indepth review that focuses on multi-sensor fusion based perception.  ...  We hope that our detailed review can help researchers to embark investigations in the area of multi-modal 3D object detection.  ...  Fuse the semantic features outputted by a segmentation model with the features of LIDAR points Same as the last one.  ... 
arXiv:2106.12735v2 fatcat:5twzbk4yhrcfzddp7zghnsivna

View suggestion for interactive segmentation of indoor scenes

Sheng Yang, Jie Xu, Kang Chen, Hongbo Fu
2017 Computational Visual Media  
Our system automatically suggests a series of camera views, in which users can conveniently specify segmentation guidance.  ...  In this way, users may focus on specifying segmentation hints instead of manually searching for desirable views of unsegmented objects, thus significantly reducing user effort.  ...  For multi-storey cases, our system provides view suggestions storey by storey.  ... 
doi:10.1007/s41095-017-0078-4 fatcat:2x4djrxn5ff6tj75mp7ns6lppe

The NEOLIX Open Dataset for Autonomous Driving [article]

Lichao Wang, Lanxin Lei, Hongli Song, Weibao Wang
2021 arXiv   pre-print
Autonomous driving vehicles rely on the co-operation of artificial intelligence, visual comput-ing, radar, monitoring equipment and GPS, whichenables computers to operate motor vehicles auto-matically  ...  s more, we are also looking forward to completing tasks such as 2D detection, object tracking, and semantic segmentation [12, 15] .  ...  The Audi Autonomous Driving Dataset [10] (A2D2) in 2020, features 2D semantic segmentation, 3D point clouds, 3D bounding boxes, and vehicle bus data.  ... 
arXiv:2011.13528v2 fatcat:ertihxfgxffpxbinqvzqauvxqm
« Previous Showing results 1 — 15 out of 4,091 results