Dense Point Diffusion for 3D Object Detection

The backbone network adopted in state-of-the-art 3D object detectors lacks a good balance between high point resolution and large receptive field, both of which are desirable for object detection on point clouds. This work proposes Dense Point Diffusion module, a novel backbone network that solves t...

Full description

Saved in:
Bibliographic Details
Main Authors: Liu, Xu, Cao, Jiayan, Bi, Qianqian, Wang, Jian, Shi, Boxin, Wei, Yichen
Format: Conference Proceeding
Language:English
Subjects:
Convolution
Feature extraction
Object detection
Quantization (signal)
Task analysis
Three-dimensional displays
Two dimensional displays
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 770
container_issue
container_start_page 762
container_title
container_volume
creator Liu, Xu
Cao, Jiayan
Bi, Qianqian
Wang, Jian
Shi, Boxin
Wei, Yichen
description The backbone network adopted in state-of-the-art 3D object detectors lacks a good balance between high point resolution and large receptive field, both of which are desirable for object detection on point clouds. This work proposes Dense Point Diffusion module, a novel backbone network that solves these issues. It adopts dilated point convolution as a building block to enlarge the receptive field and retain the point resolution at the same time. Further, a number of such layers are densely connected, giving rise to large receptive field and multi-scale feature fusion, which are effective for object detection task. Comprehensive experiments verify the efficacy of our approach. The source code 1 has been released to facilitate the reproduction of the results. 1 https://github.com/AsahiLiu/PointDetectron
doi_str_mv 10.1109/3DV50981.2020.00086
format conference_proceeding
fullrecord <record><control><sourceid>ieee_CHZPO</sourceid><recordid>TN_cdi_ieee_primary_9320330</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9320330</ieee_id><sourcerecordid>9320330</sourcerecordid><originalsourceid>FETCH-LOGICAL-i203t-65cbebcfeaa7dce570ec476e842d60e3ba40ebed390afe4481033e57a6c723943</originalsourceid><addsrcrecordid>eNotjMFKxDAUAKMguK77BXvpD7S-5KXJy1G26goL60G9Lkn6Alm0lbYe_HsDehoYhhFiK6GREtwddu8tOJKNAgUNAJC5EDfSKpIkFeGlWClt29oS0bXYzPO5NIq0JHArUXc8zFy9jHlYqi6n9D3ncajSOFXYVcdw5lg8LwXF34qr5D9m3vxzLd4eH153-_pwfHre3R_qrACX2rQxcIiJvbd95NYCR20Nk1a9AcbgNXDgHh34xFqTBMSSeROtQqdxLbZ_38zMp68pf_rp5-Sw3BHwF6olQok</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Dense Point Diffusion for 3D Object Detection</title><source>IEEE Xplore All Conference Series</source><creator>Liu, Xu ; Cao, Jiayan ; Bi, Qianqian ; Wang, Jian ; Shi, Boxin ; Wei, Yichen</creator><creatorcontrib>Liu, Xu ; Cao, Jiayan ; Bi, Qianqian ; Wang, Jian ; Shi, Boxin ; Wei, Yichen</creatorcontrib><description>The backbone network adopted in state-of-the-art 3D object detectors lacks a good balance between high point resolution and large receptive field, both of which are desirable for object detection on point clouds. This work proposes Dense Point Diffusion module, a novel backbone network that solves these issues. It adopts dilated point convolution as a building block to enlarge the receptive field and retain the point resolution at the same time. Further, a number of such layers are densely connected, giving rise to large receptive field and multi-scale feature fusion, which are effective for object detection task. Comprehensive experiments verify the efficacy of our approach. The source code 1 has been released to facilitate the reproduction of the results. 1 https://github.com/AsahiLiu/PointDetectron</description><identifier>EISSN: 2475-7888</identifier><identifier>EISBN: 1728181283</identifier><identifier>EISBN: 9781728181288</identifier><identifier>DOI: 10.1109/3DV50981.2020.00086</identifier><identifier>CODEN: IEEPAD</identifier><language>eng</language><publisher>IEEE</publisher><subject>Convolution ; Feature extraction ; Object detection ; Quantization (signal) ; Task analysis ; Three-dimensional displays ; Two dimensional displays</subject><ispartof>2020 International Conference on 3D Vision (3DV), 2020, p.762-770</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9320330$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,23930,23931,25140,27925,54555,54932</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/9320330$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Liu, Xu</creatorcontrib><creatorcontrib>Cao, Jiayan</creatorcontrib><creatorcontrib>Bi, Qianqian</creatorcontrib><creatorcontrib>Wang, Jian</creatorcontrib><creatorcontrib>Shi, Boxin</creatorcontrib><creatorcontrib>Wei, Yichen</creatorcontrib><title>Dense Point Diffusion for 3D Object Detection</title><title>2020 International Conference on 3D Vision (3DV)</title><addtitle>3DV</addtitle><description>The backbone network adopted in state-of-the-art 3D object detectors lacks a good balance between high point resolution and large receptive field, both of which are desirable for object detection on point clouds. This work proposes Dense Point Diffusion module, a novel backbone network that solves these issues. It adopts dilated point convolution as a building block to enlarge the receptive field and retain the point resolution at the same time. Further, a number of such layers are densely connected, giving rise to large receptive field and multi-scale feature fusion, which are effective for object detection task. Comprehensive experiments verify the efficacy of our approach. The source code 1 has been released to facilitate the reproduction of the results. 1 https://github.com/AsahiLiu/PointDetectron</description><subject>Convolution</subject><subject>Feature extraction</subject><subject>Object detection</subject><subject>Quantization (signal)</subject><subject>Task analysis</subject><subject>Three-dimensional displays</subject><subject>Two dimensional displays</subject><issn>2475-7888</issn><isbn>1728181283</isbn><isbn>9781728181288</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2020</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNotjMFKxDAUAKMguK77BXvpD7S-5KXJy1G26goL60G9Lkn6Alm0lbYe_HsDehoYhhFiK6GREtwddu8tOJKNAgUNAJC5EDfSKpIkFeGlWClt29oS0bXYzPO5NIq0JHArUXc8zFy9jHlYqi6n9D3ncajSOFXYVcdw5lg8LwXF34qr5D9m3vxzLd4eH153-_pwfHre3R_qrACX2rQxcIiJvbd95NYCR20Nk1a9AcbgNXDgHh34xFqTBMSSeROtQqdxLbZ_38zMp68pf_rp5-Sw3BHwF6olQok</recordid><startdate>202011</startdate><enddate>202011</enddate><creator>Liu, Xu</creator><creator>Cao, Jiayan</creator><creator>Bi, Qianqian</creator><creator>Wang, Jian</creator><creator>Shi, Boxin</creator><creator>Wei, Yichen</creator><general>IEEE</general><scope>6IE</scope><scope>6IL</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIL</scope></search><sort><creationdate>202011</creationdate><title>Dense Point Diffusion for 3D Object Detection</title><author>Liu, Xu ; Cao, Jiayan ; Bi, Qianqian ; Wang, Jian ; Shi, Boxin ; Wei, Yichen</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i203t-65cbebcfeaa7dce570ec476e842d60e3ba40ebed390afe4481033e57a6c723943</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Convolution</topic><topic>Feature extraction</topic><topic>Object detection</topic><topic>Quantization (signal)</topic><topic>Task analysis</topic><topic>Three-dimensional displays</topic><topic>Two dimensional displays</topic><toplevel>online_resources</toplevel><creatorcontrib>Liu, Xu</creatorcontrib><creatorcontrib>Cao, Jiayan</creatorcontrib><creatorcontrib>Bi, Qianqian</creatorcontrib><creatorcontrib>Wang, Jian</creatorcontrib><creatorcontrib>Shi, Boxin</creatorcontrib><creatorcontrib>Wei, Yichen</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Electronic Library Online</collection><collection>IEEE Proceedings Order Plans (POP All) 1998-Present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Liu, Xu</au><au>Cao, Jiayan</au><au>Bi, Qianqian</au><au>Wang, Jian</au><au>Shi, Boxin</au><au>Wei, Yichen</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Dense Point Diffusion for 3D Object Detection</atitle><btitle>2020 International Conference on 3D Vision (3DV)</btitle><stitle>3DV</stitle><date>2020-11</date><risdate>2020</risdate><spage>762</spage><epage>770</epage><pages>762-770</pages><eissn>2475-7888</eissn><eisbn>1728181283</eisbn><eisbn>9781728181288</eisbn><coden>IEEPAD</coden><abstract>The backbone network adopted in state-of-the-art 3D object detectors lacks a good balance between high point resolution and large receptive field, both of which are desirable for object detection on point clouds. This work proposes Dense Point Diffusion module, a novel backbone network that solves these issues. It adopts dilated point convolution as a building block to enlarge the receptive field and retain the point resolution at the same time. Further, a number of such layers are densely connected, giving rise to large receptive field and multi-scale feature fusion, which are effective for object detection task. Comprehensive experiments verify the efficacy of our approach. The source code 1 has been released to facilitate the reproduction of the results. 1 https://github.com/AsahiLiu/PointDetectron</abstract><pub>IEEE</pub><doi>10.1109/3DV50981.2020.00086</doi><tpages>9</tpages></addata></record>
fulltext fulltext_linktorsrc
identifier EISSN: 2475-7888
ispartof 2020 International Conference on 3D Vision (3DV), 2020, p.762-770
issn 2475-7888
language eng
recordid cdi_ieee_primary_9320330
source IEEE Xplore All Conference Series
subjects Convolution
Feature extraction
Object detection
Quantization (signal)
Task analysis
Three-dimensional displays
Two dimensional displays
title Dense Point Diffusion for 3D Object Detection
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-25T21%3A06%3A23IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_CHZPO&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Dense%20Point%20Diffusion%20for%203D%20Object%20Detection&rft.btitle=2020%20International%20Conference%20on%203D%20Vision%20(3DV)&rft.au=Liu,%20Xu&rft.date=2020-11&rft.spage=762&rft.epage=770&rft.pages=762-770&rft.eissn=2475-7888&rft.coden=IEEPAD&rft_id=info:doi/10.1109/3DV50981.2020.00086&rft.eisbn=1728181283&rft.eisbn_list=9781728181288&rft_dat=%3Cieee_CHZPO%3E9320330%3C/ieee_CHZPO%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i203t-65cbebcfeaa7dce570ec476e842d60e3ba40ebed390afe4481033e57a6c723943%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=9320330&rfr_iscdi=true