Research on Data Augmentation and Object Detection for Small Sample Data

Zhang, Can

Research on Data Augmentation and Object Detection for Small Sample Data

Zhang, Can

Permalink

Publication Type:: Thesis
Issue Date:: 2023

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download thesisAdobe PDF (13.89 MB)

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Zhang, Can
dc.date.accessioned	2023-07-17T00:06:57Z
dc.date.available	2023-07-17T00:06:57Z
dc.date.issued	2023
dc.identifier.uri	http://hdl.handle.net/10453/171524
dc.description	University of Technology Sydney. Faculty of Engineering and Information Technology.	en_US.UTF-8
dc.description.abstract	Aiming at the problems of high time cost of training data collection and large sample scale span in deep learning tasks, this thesis first presents a data augmentation method based on unsupervised style transfer, which can effectively increase the data in a small sample dataset; Second, for different data augmentation methods, an universal assessment system is constructed, where various data expansion approaches are comprehensively evaluated in terms of classification accuracy and data variety; Then, for the problem of small target detection in small sample data set, the Improved Cascade RCNN model for small target detection is proposed; Finally, the above methods are applied to the example of intelligent garbage detection system, which considerably increases the overall garbage detection accuracy.	en_US.UTF-8
dc.format	Thesis (PhD)
dc.language.iso	en_US	en_US.UTF-8
dc.relation	https://opus.lib.uts.edu.au/bitstream/10453/171524/1/thesis.pdf
dc.rights	info:eu-repo/semantics/openAccess
dc.rights	The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.
dc.rights	© 2023 Can Zhang
dc.rights	au.edu.uts.lib/nph
dc.title	Research on Data Augmentation and Object Detection for Small Sample Data	en_US.UTF-8
dc.type	Thesis
utslib.copyright.status	open_access	*

Abstract:

Aiming at the problems of high time cost of training data collection and large sample scale span in deep learning tasks, this thesis first presents a data augmentation method based on unsupervised style transfer, which can effectively increase the data in a small sample dataset; Second, for different data augmentation methods, an universal assessment system is constructed, where various data expansion approaches are comprehensively evaluated in terms of classification accuracy and data variety; Then, for the problem of small target detection in small sample data set, the Improved Cascade RCNN model for small target detection is proposed; Finally, the above methods are applied to the example of intelligent garbage detection system, which considerably increases the overall garbage detection accuracy.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/171524