Full metadata record
DC FieldValueLanguage
dc.contributorDepartment of Mechanical Engineeringen_US
dc.creatorFang, Hong-
dc.identifier.urihttps://theses.lib.polyu.edu.hk/handle/200/11481-
dc.languageEnglishen_US
dc.publisherHong Kong Polytechnic Universityen_US
dc.rightsAll rights reserveden_US
dc.titleOne stage grasp detection with transfer learningen_US
dcterms.abstractGrasp detection plays an important role in robotic application. In recent years, due to successful utilization of deep learning in computer vision, natural language processing and autonomous driving, grasp detection with deep learning has inspired a great deal of researchers' attention. Compared with arduous and time-consuming traditional strategies applied in grasp detection, deep learning can significantly enhance the accuracy and generalization ability of robotic grasping aiming for novel objects. In this work, we propose a novel convolutional neural network architecture based on VGG-16 and follow the design ideas of YOLO to realize grasps detection aiming for five-dimensional representation. In order to adapt our training strategy for oriented bounding box in grasp detection while increasing network's regression accuracy and training speed, we propose a new angle matching mechanism to replace Jaccard Index Matching. Meanwhile, transfer learning is also used in our strategy to enhance network's ability of feature extraction and end-to-end training speed. Five-fold cross validation indicates our proposed algorithm has satisfied performance in predicting objects' grasp configurations, which achieves an accuracy of 78.7% and 77.3% in image-wise split and object-wise split respectively after only training 5 epochs.en_US
dcterms.extent53 pages : color illustrationsen_US
dcterms.isPartOfPolyU Electronic Thesesen_US
dcterms.issued2021en_US
dcterms.educationalLevelM.Sc.en_US
dcterms.educationalLevelAll Masteren_US
dcterms.LCSHRoboticsen_US
dcterms.LCSHHong Kong Polytechnic University -- Dissertationsen_US
dcterms.accessRightsrestricted accessen_US

Files in This Item:
File Description SizeFormat 
5927.pdfFor All Users (off-campus access for PolyU Staff & Students only)3.75 MBAdobe PDFView/Open


Copyright Undertaking

As a bona fide Library user, I declare that:

  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.

By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.

Show simple item record

Please use this identifier to cite or link to this item: https://theses.lib.polyu.edu.hk/handle/200/11481