Domain adaption for few-shot learning in image classification

Machine learning has been widely used in various fields and successfully addresses many image classification problems in the presence of sufficient samples, but performs poorly in the absence of samples. Few-shot learning is an innovative approach to solving this problem. In this article, I con...

Full description

Saved in:
Bibliographic Details
Main Author: Pan, Yifei
Other Authors: Mao Kezhi
Format: Thesis-Master by Coursework
Language:English
Published: Nanyang Technological University 2023
Subjects:
Online Access:https://hdl.handle.net/10356/164914
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-164914
record_format dspace
spelling sg-ntu-dr.10356-1649142023-02-27T01:11:04Z Domain adaption for few-shot learning in image classification Pan, Yifei Mao Kezhi School of Electrical and Electronic Engineering EKZMao@ntu.edu.sg Engineering::Electrical and electronic engineering Machine learning has been widely used in various fields and successfully addresses many image classification problems in the presence of sufficient samples, but performs poorly in the absence of samples. Few-shot learning is an innovative approach to solving this problem. In this article, I conduct a comprehensive survey and categorize few-shot learning algorithms from three perspectives: (i) initialization based methods; (ii) distance metric learning based methods; and (iii) hallucination based methods. I further study and evaluate (1) Fine-tuning; (2) Model-Agnostic Meta-Learning; (3) Prototypical Networks; (4) Relation Networks using the Omniglot and miniImageNet datasets. Master of Science (Signal Processing) 2023-02-27T01:11:04Z 2023-02-27T01:11:04Z 2023 Thesis-Master by Coursework Pan, Y. (2023). Domain adaption for few-shot learning in image classification. Master's thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/164914 https://hdl.handle.net/10356/164914 en ISM-DISS-03090 application/pdf Nanyang Technological University
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Electrical and electronic engineering
spellingShingle Engineering::Electrical and electronic engineering
Pan, Yifei
Domain adaption for few-shot learning in image classification
description Machine learning has been widely used in various fields and successfully addresses many image classification problems in the presence of sufficient samples, but performs poorly in the absence of samples. Few-shot learning is an innovative approach to solving this problem. In this article, I conduct a comprehensive survey and categorize few-shot learning algorithms from three perspectives: (i) initialization based methods; (ii) distance metric learning based methods; and (iii) hallucination based methods. I further study and evaluate (1) Fine-tuning; (2) Model-Agnostic Meta-Learning; (3) Prototypical Networks; (4) Relation Networks using the Omniglot and miniImageNet datasets.
author2 Mao Kezhi
author_facet Mao Kezhi
Pan, Yifei
format Thesis-Master by Coursework
author Pan, Yifei
author_sort Pan, Yifei
title Domain adaption for few-shot learning in image classification
title_short Domain adaption for few-shot learning in image classification
title_full Domain adaption for few-shot learning in image classification
title_fullStr Domain adaption for few-shot learning in image classification
title_full_unstemmed Domain adaption for few-shot learning in image classification
title_sort domain adaption for few-shot learning in image classification
publisher Nanyang Technological University
publishDate 2023
url https://hdl.handle.net/10356/164914
_version_ 1759058829584105472