Exploiting Sparsity in Automotive Radar Object Detection Networks
Having precise perception of the environment is crucial for ensuring the secure and reliable functioning of autonomous driving systems. Radar object detection networks are one fundamental part of such systems. CNN-based object detectors showed good performance in this context, but they require large...
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | Lippke, Marius Quach, Maurice Braun, Sascha Köhler, Daniel Ulrich, Michael Bischoff, Bastian Tan, Wei Yap |
| description | Having precise perception of the environment is crucial for ensuring the
secure and reliable functioning of autonomous driving systems. Radar object
detection networks are one fundamental part of such systems. CNN-based object
detectors showed good performance in this context, but they require large
compute resources. This paper investigates sparse convolutional object
detection networks, which combine powerful grid-based detection with low
compute resources. We investigate radar specific challenges and propose sparse
kernel point pillars (SKPP) and dual voxel point convolutions (DVPC) as
remedies for the grid rendering and sparse backbone architectures. We evaluate
our SKPP-DPVCN architecture on nuScenes, which outperforms the baseline by
5.89% and the previous state of the art by 4.19% in Car AP4.0. Moreover,
SKPP-DPVCN reduces the average scale error (ASE) by 21.41% over the baseline. |
| doi_str_mv | 10.48550/arxiv.2308.07748 |
| format | Article |
| fullrecord | <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2308_07748</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2308_07748</sourcerecordid><originalsourceid>FETCH-LOGICAL-a678-b292542b0c5c565f3d0143e2cf5a2dbf925bf776bf97da04f442400d1c4ae80c3</originalsourceid><addsrcrecordid>eNotj8tOwzAUBb1hgQofwAr_QMKNY8fuMirlIVWtBN1H16_K0MaRY0r794TCao50pJGGkLsKSq6EgAdMp3AsWQ2qBCm5uibt8jTsY8ih39H3AdMY8pmGnrZfOR5iDkdH39Biohv94Uymjy5PCLGna5e_Y_ocb8iVx_3obv85I9un5XbxUqw2z6-LdlVgI1Wh2ZwJzjQYYUQjfG2h4rVjxgtkVvvp1V7KZlrSInDPOeMAtjIcnQJTz8j9n_bS0A0pHDCdu9-W7tJS_wCjUUSZ</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Exploiting Sparsity in Automotive Radar Object Detection Networks</title><source>arXiv.org</source><creator>Lippke, Marius ; Quach, Maurice ; Braun, Sascha ; Köhler, Daniel ; Ulrich, Michael ; Bischoff, Bastian ; Tan, Wei Yap</creator><creatorcontrib>Lippke, Marius ; Quach, Maurice ; Braun, Sascha ; Köhler, Daniel ; Ulrich, Michael ; Bischoff, Bastian ; Tan, Wei Yap</creatorcontrib><description>Having precise perception of the environment is crucial for ensuring the
secure and reliable functioning of autonomous driving systems. Radar object
detection networks are one fundamental part of such systems. CNN-based object
detectors showed good performance in this context, but they require large
compute resources. This paper investigates sparse convolutional object
detection networks, which combine powerful grid-based detection with low
compute resources. We investigate radar specific challenges and propose sparse
kernel point pillars (SKPP) and dual voxel point convolutions (DVPC) as
remedies for the grid rendering and sparse backbone architectures. We evaluate
our SKPP-DPVCN architecture on nuScenes, which outperforms the baseline by
5.89% and the previous state of the art by 4.19% in Car AP4.0. Moreover,
SKPP-DPVCN reduces the average scale error (ASE) by 21.41% over the baseline.</description><identifier>DOI: 10.48550/arxiv.2308.07748</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Learning ; Computer Science - Robotics</subject><creationdate>2023-08</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2308.07748$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2308.07748$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Lippke, Marius</creatorcontrib><creatorcontrib>Quach, Maurice</creatorcontrib><creatorcontrib>Braun, Sascha</creatorcontrib><creatorcontrib>Köhler, Daniel</creatorcontrib><creatorcontrib>Ulrich, Michael</creatorcontrib><creatorcontrib>Bischoff, Bastian</creatorcontrib><creatorcontrib>Tan, Wei Yap</creatorcontrib><title>Exploiting Sparsity in Automotive Radar Object Detection Networks</title><description>Having precise perception of the environment is crucial for ensuring the
secure and reliable functioning of autonomous driving systems. Radar object
detection networks are one fundamental part of such systems. CNN-based object
detectors showed good performance in this context, but they require large
compute resources. This paper investigates sparse convolutional object
detection networks, which combine powerful grid-based detection with low
compute resources. We investigate radar specific challenges and propose sparse
kernel point pillars (SKPP) and dual voxel point convolutions (DVPC) as
remedies for the grid rendering and sparse backbone architectures. We evaluate
our SKPP-DPVCN architecture on nuScenes, which outperforms the baseline by
5.89% and the previous state of the art by 4.19% in Car AP4.0. Moreover,
SKPP-DPVCN reduces the average scale error (ASE) by 21.41% over the baseline.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Computer Vision and Pattern Recognition</subject><subject>Computer Science - Learning</subject><subject>Computer Science - Robotics</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj8tOwzAUBb1hgQofwAr_QMKNY8fuMirlIVWtBN1H16_K0MaRY0r794TCao50pJGGkLsKSq6EgAdMp3AsWQ2qBCm5uibt8jTsY8ih39H3AdMY8pmGnrZfOR5iDkdH39Biohv94Uymjy5PCLGna5e_Y_ocb8iVx_3obv85I9un5XbxUqw2z6-LdlVgI1Wh2ZwJzjQYYUQjfG2h4rVjxgtkVvvp1V7KZlrSInDPOeMAtjIcnQJTz8j9n_bS0A0pHDCdu9-W7tJS_wCjUUSZ</recordid><startdate>20230815</startdate><enddate>20230815</enddate><creator>Lippke, Marius</creator><creator>Quach, Maurice</creator><creator>Braun, Sascha</creator><creator>Köhler, Daniel</creator><creator>Ulrich, Michael</creator><creator>Bischoff, Bastian</creator><creator>Tan, Wei Yap</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20230815</creationdate><title>Exploiting Sparsity in Automotive Radar Object Detection Networks</title><author>Lippke, Marius ; Quach, Maurice ; Braun, Sascha ; Köhler, Daniel ; Ulrich, Michael ; Bischoff, Bastian ; Tan, Wei Yap</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a678-b292542b0c5c565f3d0143e2cf5a2dbf925bf776bf97da04f442400d1c4ae80c3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Computer Vision and Pattern Recognition</topic><topic>Computer Science - Learning</topic><topic>Computer Science - Robotics</topic><toplevel>online_resources</toplevel><creatorcontrib>Lippke, Marius</creatorcontrib><creatorcontrib>Quach, Maurice</creatorcontrib><creatorcontrib>Braun, Sascha</creatorcontrib><creatorcontrib>Köhler, Daniel</creatorcontrib><creatorcontrib>Ulrich, Michael</creatorcontrib><creatorcontrib>Bischoff, Bastian</creatorcontrib><creatorcontrib>Tan, Wei Yap</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Lippke, Marius</au><au>Quach, Maurice</au><au>Braun, Sascha</au><au>Köhler, Daniel</au><au>Ulrich, Michael</au><au>Bischoff, Bastian</au><au>Tan, Wei Yap</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Exploiting Sparsity in Automotive Radar Object Detection Networks</atitle><date>2023-08-15</date><risdate>2023</risdate><abstract>Having precise perception of the environment is crucial for ensuring the
secure and reliable functioning of autonomous driving systems. Radar object
detection networks are one fundamental part of such systems. CNN-based object
detectors showed good performance in this context, but they require large
compute resources. This paper investigates sparse convolutional object
detection networks, which combine powerful grid-based detection with low
compute resources. We investigate radar specific challenges and propose sparse
kernel point pillars (SKPP) and dual voxel point convolutions (DVPC) as
remedies for the grid rendering and sparse backbone architectures. We evaluate
our SKPP-DPVCN architecture on nuScenes, which outperforms the baseline by
5.89% and the previous state of the art by 4.19% in Car AP4.0. Moreover,
SKPP-DPVCN reduces the average scale error (ASE) by 21.41% over the baseline.</abstract><doi>10.48550/arxiv.2308.07748</doi><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | DOI: 10.48550/arxiv.2308.07748 |
| ispartof | |
| issn | |
| language | eng |
| recordid | cdi_arxiv_primary_2308_07748 |
| source | arXiv.org |
| subjects | Computer Science - Artificial Intelligence Computer Science - Computer Vision and Pattern Recognition Computer Science - Learning Computer Science - Robotics |
| title | Exploiting Sparsity in Automotive Radar Object Detection Networks |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-11T12%3A20%3A55IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Exploiting%20Sparsity%20in%20Automotive%20Radar%20Object%20Detection%20Networks&rft.au=Lippke,%20Marius&rft.date=2023-08-15&rft_id=info:doi/10.48550/arxiv.2308.07748&rft_dat=%3Carxiv_GOX%3E2308_07748%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |