A Principal Component Analysis-Based Feature Optimization Network for Few-Shot Fine-Grained Image Classification

Feature map reconstruction networks (FRN) have demonstrated significant potential by leveraging feature reconstruction. However, the typical process of FRN gives rise to two notable issues. First, FRN exhibits high sensitivity to noise, particularly ambient noise, which can lead to substantial recon...

Full description

Saved in:
Bibliographic Details
Main Authors: Meijia Wang, Boyuan Zheng, Guochao Wang, Junpo Yang, Jin Lu, Weichuan Zhang
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:Mathematics
Subjects:
Online Access:https://www.mdpi.com/2227-7390/13/7/1098
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849730292456095744
author Meijia Wang
Boyuan Zheng
Guochao Wang
Junpo Yang
Jin Lu
Weichuan Zhang
author_facet Meijia Wang
Boyuan Zheng
Guochao Wang
Junpo Yang
Jin Lu
Weichuan Zhang
author_sort Meijia Wang
collection DOAJ
description Feature map reconstruction networks (FRN) have demonstrated significant potential by leveraging feature reconstruction. However, the typical process of FRN gives rise to two notable issues. First, FRN exhibits high sensitivity to noise, particularly ambient noise, which can lead to substantial reconstruction errors and hinder the network’s ability to extract meaningful features. Second, FRN is particularly vulnerable to changes in data distribution. Owing to the fine-grained nature of the training data, the model is highly susceptible to overfitting, which may compromise its ability to extract effective feature representations when confronted with new classes. To address these challenges, this paper proposes a novel main feature selection module (MFSM), which suppresses feature noise interference and enhances the discriminative capacity of feature representations through principal component analysis (PCA). Extensive experiments validate the effectiveness of MFSM, revealing substantial improvements in classification accuracy for few-shot fine-grained image classification (FSFGIC) tasks.
format Article
id doaj-art-ebf7764ee7d64fd6a93d984e7f24d008
institution DOAJ
issn 2227-7390
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series Mathematics
spelling doaj-art-ebf7764ee7d64fd6a93d984e7f24d0082025-08-20T03:08:55ZengMDPI AGMathematics2227-73902025-03-01137109810.3390/math13071098A Principal Component Analysis-Based Feature Optimization Network for Few-Shot Fine-Grained Image ClassificationMeijia Wang0Boyuan Zheng1Guochao Wang2Junpo Yang3Jin Lu4Weichuan Zhang5School of Electronic Information and Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an 710021, ChinaSchool of Electronic Information and Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an 710021, ChinaSchool of Electronic Information and Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an 710021, ChinaSchool of Electronic Information and Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an 710021, ChinaSchool of Electronic Information and Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an 710021, ChinaSchool of Electronic Information and Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an 710021, ChinaFeature map reconstruction networks (FRN) have demonstrated significant potential by leveraging feature reconstruction. However, the typical process of FRN gives rise to two notable issues. First, FRN exhibits high sensitivity to noise, particularly ambient noise, which can lead to substantial reconstruction errors and hinder the network’s ability to extract meaningful features. Second, FRN is particularly vulnerable to changes in data distribution. Owing to the fine-grained nature of the training data, the model is highly susceptible to overfitting, which may compromise its ability to extract effective feature representations when confronted with new classes. To address these challenges, this paper proposes a novel main feature selection module (MFSM), which suppresses feature noise interference and enhances the discriminative capacity of feature representations through principal component analysis (PCA). Extensive experiments validate the effectiveness of MFSM, revealing substantial improvements in classification accuracy for few-shot fine-grained image classification (FSFGIC) tasks.https://www.mdpi.com/2227-7390/13/7/1098few-shot fine-grained image classificationprincipal component analysisprincipal component analysis-based feature optimization network
spellingShingle Meijia Wang
Boyuan Zheng
Guochao Wang
Junpo Yang
Jin Lu
Weichuan Zhang
A Principal Component Analysis-Based Feature Optimization Network for Few-Shot Fine-Grained Image Classification
Mathematics
few-shot fine-grained image classification
principal component analysis
principal component analysis-based feature optimization network
title A Principal Component Analysis-Based Feature Optimization Network for Few-Shot Fine-Grained Image Classification
title_full A Principal Component Analysis-Based Feature Optimization Network for Few-Shot Fine-Grained Image Classification
title_fullStr A Principal Component Analysis-Based Feature Optimization Network for Few-Shot Fine-Grained Image Classification
title_full_unstemmed A Principal Component Analysis-Based Feature Optimization Network for Few-Shot Fine-Grained Image Classification
title_short A Principal Component Analysis-Based Feature Optimization Network for Few-Shot Fine-Grained Image Classification
title_sort principal component analysis based feature optimization network for few shot fine grained image classification
topic few-shot fine-grained image classification
principal component analysis
principal component analysis-based feature optimization network
url https://www.mdpi.com/2227-7390/13/7/1098
work_keys_str_mv AT meijiawang aprincipalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT boyuanzheng aprincipalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT guochaowang aprincipalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT junpoyang aprincipalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT jinlu aprincipalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT weichuanzhang aprincipalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT meijiawang principalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT boyuanzheng principalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT guochaowang principalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT junpoyang principalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT jinlu principalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification
AT weichuanzhang principalcomponentanalysisbasedfeatureoptimizationnetworkforfewshotfinegrainedimageclassification