VMGNet: A Low Computational Complexity Robotic Grasping Network Based on VMamba with Multi-Scale Feature Fusion
While deep learning-based robotic grasping technology has demonstrated strong adaptability, its computational complexity has also significantly increased, making it unsuitable for scenarios with high real-time requirements. Therefore, we propose a low computational complexity and high accuracy model...
Gespeichert in:
Veröffentlicht in: | arXiv.org 2024-11 |
---|---|
Hauptverfasser: | , , , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | |
container_start_page | |
container_title | arXiv.org |
container_volume | |
creator | Jin, Yuhao Gao, Qizhong Zhu, Xiaohui Yue, Yong Eng Gee Lim Chen, Yuqing Wong, Prudence Chu, Yijie |
description | While deep learning-based robotic grasping technology has demonstrated strong adaptability, its computational complexity has also significantly increased, making it unsuitable for scenarios with high real-time requirements. Therefore, we propose a low computational complexity and high accuracy model named VMGNet for robotic grasping. For the first time, we introduce the Visual State Space into the robotic grasping field to achieve linear computational complexity, thereby greatly reducing the model's computational cost. Meanwhile, to improve the accuracy of the model, we propose an efficient and lightweight multi-scale feature fusion module, named Fusion Bridge Module, to extract and fuse information at different scales. We also present a new loss function calculation method to enhance the importance differences between subtasks, improving the model's fitting ability. Experiments show that VMGNet has only 8.7G Floating Point Operations and an inference time of 8.1 ms on our devices. VMGNet also achieved state-of-the-art performance on the Cornell and Jacquard public datasets. To validate VMGNet's effectiveness in practical applications, we conducted real grasping experiments in multi-object scenarios, and VMGNet achieved an excellent performance with a 94.4% success rate in real-world grasping tasks. The video for the real-world robotic grasping experiments is available at https://youtu.be/S-QHBtbmLc4. |
format | Article |
fullrecord | <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_3130965953</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>3130965953</sourcerecordid><originalsourceid>FETCH-proquest_journals_31309659533</originalsourceid><addsrcrecordid>eNqNjcGKwjAURYMgjKj_8MB1oSa2o-4cGZ2FdaHiVl41ajT21eSFjn9vET_A1eHCPfc2REsq1Y-GAym_RNf7SxzHMv2WSaJagrbZfKl5DBNYUAVTupWBkQ0VaF_J6n_DD1hRTmz2MHfoS1OcoJYqclf4Qa8PQAVsM7zlCJXhM2TBsonWe7QaZho5uJrB16sd0Tyi9br7Zlv0Zr-b6V9UOroH7Xl3oeDqc79TfRWP0mSUKPVZ6wnPPUlb</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>3130965953</pqid></control><display><type>article</type><title>VMGNet: A Low Computational Complexity Robotic Grasping Network Based on VMamba with Multi-Scale Feature Fusion</title><source>Free E- Journals</source><creator>Jin, Yuhao ; Gao, Qizhong ; Zhu, Xiaohui ; Yue, Yong ; Eng Gee Lim ; Chen, Yuqing ; Wong, Prudence ; Chu, Yijie</creator><creatorcontrib>Jin, Yuhao ; Gao, Qizhong ; Zhu, Xiaohui ; Yue, Yong ; Eng Gee Lim ; Chen, Yuqing ; Wong, Prudence ; Chu, Yijie</creatorcontrib><description>While deep learning-based robotic grasping technology has demonstrated strong adaptability, its computational complexity has also significantly increased, making it unsuitable for scenarios with high real-time requirements. Therefore, we propose a low computational complexity and high accuracy model named VMGNet for robotic grasping. For the first time, we introduce the Visual State Space into the robotic grasping field to achieve linear computational complexity, thereby greatly reducing the model's computational cost. Meanwhile, to improve the accuracy of the model, we propose an efficient and lightweight multi-scale feature fusion module, named Fusion Bridge Module, to extract and fuse information at different scales. We also present a new loss function calculation method to enhance the importance differences between subtasks, improving the model's fitting ability. Experiments show that VMGNet has only 8.7G Floating Point Operations and an inference time of 8.1 ms on our devices. VMGNet also achieved state-of-the-art performance on the Cornell and Jacquard public datasets. To validate VMGNet's effectiveness in practical applications, we conducted real grasping experiments in multi-object scenarios, and VMGNet achieved an excellent performance with a 94.4% success rate in real-world grasping tasks. The video for the real-world robotic grasping experiments is available at https://youtu.be/S-QHBtbmLc4.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Accuracy ; Complexity ; Computing costs ; Floating point arithmetic ; Grasping (robotics) ; Modules ; Real time ; Robotics ; Visual fields</subject><ispartof>arXiv.org, 2024-11</ispartof><rights>2024. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</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>780,784</link.rule.ids></links><search><creatorcontrib>Jin, Yuhao</creatorcontrib><creatorcontrib>Gao, Qizhong</creatorcontrib><creatorcontrib>Zhu, Xiaohui</creatorcontrib><creatorcontrib>Yue, Yong</creatorcontrib><creatorcontrib>Eng Gee Lim</creatorcontrib><creatorcontrib>Chen, Yuqing</creatorcontrib><creatorcontrib>Wong, Prudence</creatorcontrib><creatorcontrib>Chu, Yijie</creatorcontrib><title>VMGNet: A Low Computational Complexity Robotic Grasping Network Based on VMamba with Multi-Scale Feature Fusion</title><title>arXiv.org</title><description>While deep learning-based robotic grasping technology has demonstrated strong adaptability, its computational complexity has also significantly increased, making it unsuitable for scenarios with high real-time requirements. Therefore, we propose a low computational complexity and high accuracy model named VMGNet for robotic grasping. For the first time, we introduce the Visual State Space into the robotic grasping field to achieve linear computational complexity, thereby greatly reducing the model's computational cost. Meanwhile, to improve the accuracy of the model, we propose an efficient and lightweight multi-scale feature fusion module, named Fusion Bridge Module, to extract and fuse information at different scales. We also present a new loss function calculation method to enhance the importance differences between subtasks, improving the model's fitting ability. Experiments show that VMGNet has only 8.7G Floating Point Operations and an inference time of 8.1 ms on our devices. VMGNet also achieved state-of-the-art performance on the Cornell and Jacquard public datasets. To validate VMGNet's effectiveness in practical applications, we conducted real grasping experiments in multi-object scenarios, and VMGNet achieved an excellent performance with a 94.4% success rate in real-world grasping tasks. The video for the real-world robotic grasping experiments is available at https://youtu.be/S-QHBtbmLc4.</description><subject>Accuracy</subject><subject>Complexity</subject><subject>Computing costs</subject><subject>Floating point arithmetic</subject><subject>Grasping (robotics)</subject><subject>Modules</subject><subject>Real time</subject><subject>Robotics</subject><subject>Visual fields</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNjcGKwjAURYMgjKj_8MB1oSa2o-4cGZ2FdaHiVl41ajT21eSFjn9vET_A1eHCPfc2REsq1Y-GAym_RNf7SxzHMv2WSaJagrbZfKl5DBNYUAVTupWBkQ0VaF_J6n_DD1hRTmz2MHfoS1OcoJYqclf4Qa8PQAVsM7zlCJXhM2TBsonWe7QaZho5uJrB16sd0Tyi9br7Zlv0Zr-b6V9UOroH7Xl3oeDqc79TfRWP0mSUKPVZ6wnPPUlb</recordid><startdate>20241119</startdate><enddate>20241119</enddate><creator>Jin, Yuhao</creator><creator>Gao, Qizhong</creator><creator>Zhu, Xiaohui</creator><creator>Yue, Yong</creator><creator>Eng Gee Lim</creator><creator>Chen, Yuqing</creator><creator>Wong, Prudence</creator><creator>Chu, Yijie</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20241119</creationdate><title>VMGNet: A Low Computational Complexity Robotic Grasping Network Based on VMamba with Multi-Scale Feature Fusion</title><author>Jin, Yuhao ; Gao, Qizhong ; Zhu, Xiaohui ; Yue, Yong ; Eng Gee Lim ; Chen, Yuqing ; Wong, Prudence ; Chu, Yijie</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_31309659533</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Accuracy</topic><topic>Complexity</topic><topic>Computing costs</topic><topic>Floating point arithmetic</topic><topic>Grasping (robotics)</topic><topic>Modules</topic><topic>Real time</topic><topic>Robotics</topic><topic>Visual fields</topic><toplevel>online_resources</toplevel><creatorcontrib>Jin, Yuhao</creatorcontrib><creatorcontrib>Gao, Qizhong</creatorcontrib><creatorcontrib>Zhu, Xiaohui</creatorcontrib><creatorcontrib>Yue, Yong</creatorcontrib><creatorcontrib>Eng Gee Lim</creatorcontrib><creatorcontrib>Chen, Yuqing</creatorcontrib><creatorcontrib>Wong, Prudence</creatorcontrib><creatorcontrib>Chu, Yijie</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science & Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Jin, Yuhao</au><au>Gao, Qizhong</au><au>Zhu, Xiaohui</au><au>Yue, Yong</au><au>Eng Gee Lim</au><au>Chen, Yuqing</au><au>Wong, Prudence</au><au>Chu, Yijie</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>VMGNet: A Low Computational Complexity Robotic Grasping Network Based on VMamba with Multi-Scale Feature Fusion</atitle><jtitle>arXiv.org</jtitle><date>2024-11-19</date><risdate>2024</risdate><eissn>2331-8422</eissn><abstract>While deep learning-based robotic grasping technology has demonstrated strong adaptability, its computational complexity has also significantly increased, making it unsuitable for scenarios with high real-time requirements. Therefore, we propose a low computational complexity and high accuracy model named VMGNet for robotic grasping. For the first time, we introduce the Visual State Space into the robotic grasping field to achieve linear computational complexity, thereby greatly reducing the model's computational cost. Meanwhile, to improve the accuracy of the model, we propose an efficient and lightweight multi-scale feature fusion module, named Fusion Bridge Module, to extract and fuse information at different scales. We also present a new loss function calculation method to enhance the importance differences between subtasks, improving the model's fitting ability. Experiments show that VMGNet has only 8.7G Floating Point Operations and an inference time of 8.1 ms on our devices. VMGNet also achieved state-of-the-art performance on the Cornell and Jacquard public datasets. To validate VMGNet's effectiveness in practical applications, we conducted real grasping experiments in multi-object scenarios, and VMGNet achieved an excellent performance with a 94.4% success rate in real-world grasping tasks. The video for the real-world robotic grasping experiments is available at https://youtu.be/S-QHBtbmLc4.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | EISSN: 2331-8422 |
ispartof | arXiv.org, 2024-11 |
issn | 2331-8422 |
language | eng |
recordid | cdi_proquest_journals_3130965953 |
source | Free E- Journals |
subjects | Accuracy Complexity Computing costs Floating point arithmetic Grasping (robotics) Modules Real time Robotics Visual fields |
title | VMGNet: A Low Computational Complexity Robotic Grasping Network Based on VMamba with Multi-Scale Feature Fusion |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-05T00%3A30%3A29IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=VMGNet:%20A%20Low%20Computational%20Complexity%20Robotic%20Grasping%20Network%20Based%20on%20VMamba%20with%20Multi-Scale%20Feature%20Fusion&rft.jtitle=arXiv.org&rft.au=Jin,%20Yuhao&rft.date=2024-11-19&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E3130965953%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=3130965953&rft_id=info:pmid/&rfr_iscdi=true |