Few-Shot Incremental Learning With Context-Aware Spatial Enhancement for Image Recognition

Few-shot incremental learning (FSIL) refers to the ability of a model to learn new concepts from a limited number of labeled examples and gradually recognize novel categories with minimal supervision while retaining previously learned knowledge to prevent forgetting. To address the key challenges in...

Full description

Saved in:
Bibliographic Details
Main Authors: Heng Wu, Ze Yang, Zijun Zheng, Haiyang Wang, Wansong Wang
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11036738/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849426023169392640
author Heng Wu
Ze Yang
Zijun Zheng
Haiyang Wang
Wansong Wang
author_facet Heng Wu
Ze Yang
Zijun Zheng
Haiyang Wang
Wansong Wang
author_sort Heng Wu
collection DOAJ
description Few-shot incremental learning (FSIL) refers to the ability of a model to learn new concepts from a limited number of labeled examples and gradually recognize novel categories with minimal supervision while retaining previously learned knowledge to prevent forgetting. To address the key challenges in FSIL, this paper proposes a novel Context-Aware Spatial Enhancement (CASE) framework, which improves feature representations by jointly leveraging global and local spatial information. Specifically, the global spatial enhancement module captures long-range dependencies to enrich semantic context, while the local spatial enhancement module applies Gaussian filtering to refine fine-grained details and suppress background noise. Additionally, a multi-head interaction block is introduced to model intricate relationships between spatial regions, effectively bridging global and local perspectives for robust and context-aware feature learning. CASE also incorporates prior knowledge from base categories to enhance adaptability to novel classes. This comprehensive design not only reduces background interference and highlights salient features but also improves generalization and stability in few-shot learning scenarios. Extensive experiments on benchmark datasets validate the effectiveness of the proposed approach, demonstrating competitive performance along with superior scalability and generalization capability in FSIL tasks.
format Article
id doaj-art-a6b6ef122e9f4e24a9ebcdb0c1e87576
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-a6b6ef122e9f4e24a9ebcdb0c1e875762025-08-20T03:29:34ZengIEEEIEEE Access2169-35362025-01-011311056911058310.1109/ACCESS.2025.357976611036738Few-Shot Incremental Learning With Context-Aware Spatial Enhancement for Image RecognitionHeng Wu0https://orcid.org/0000-0002-4632-0339Ze Yang1Zijun Zheng2Haiyang Wang3Wansong Wang4https://orcid.org/0000-0003-2855-8228College of Information Engineering, Hangzhou Vocational and Technical College, Hangzhou, ChinaChina Mobile Information System Integration Company Ltd., Beijing, ChinaCollege of Sciences, China Jiliang University, Hangzhou, ChinaInstitute of Network Technology (Yantai), Yantai, ChinaDepartment of Rehabilitation Medicine, The First Affiliated Hospital, Jiangxi Medical College, Nanchang University, Nanchang, ChinaFew-shot incremental learning (FSIL) refers to the ability of a model to learn new concepts from a limited number of labeled examples and gradually recognize novel categories with minimal supervision while retaining previously learned knowledge to prevent forgetting. To address the key challenges in FSIL, this paper proposes a novel Context-Aware Spatial Enhancement (CASE) framework, which improves feature representations by jointly leveraging global and local spatial information. Specifically, the global spatial enhancement module captures long-range dependencies to enrich semantic context, while the local spatial enhancement module applies Gaussian filtering to refine fine-grained details and suppress background noise. Additionally, a multi-head interaction block is introduced to model intricate relationships between spatial regions, effectively bridging global and local perspectives for robust and context-aware feature learning. CASE also incorporates prior knowledge from base categories to enhance adaptability to novel classes. This comprehensive design not only reduces background interference and highlights salient features but also improves generalization and stability in few-shot learning scenarios. Extensive experiments on benchmark datasets validate the effectiveness of the proposed approach, demonstrating competitive performance along with superior scalability and generalization capability in FSIL tasks.https://ieeexplore.ieee.org/document/11036738/Few-shot incremental learningimage recognitioncontext-aware spatial enhancement
spellingShingle Heng Wu
Ze Yang
Zijun Zheng
Haiyang Wang
Wansong Wang
Few-Shot Incremental Learning With Context-Aware Spatial Enhancement for Image Recognition
IEEE Access
Few-shot incremental learning
image recognition
context-aware spatial enhancement
title Few-Shot Incremental Learning With Context-Aware Spatial Enhancement for Image Recognition
title_full Few-Shot Incremental Learning With Context-Aware Spatial Enhancement for Image Recognition
title_fullStr Few-Shot Incremental Learning With Context-Aware Spatial Enhancement for Image Recognition
title_full_unstemmed Few-Shot Incremental Learning With Context-Aware Spatial Enhancement for Image Recognition
title_short Few-Shot Incremental Learning With Context-Aware Spatial Enhancement for Image Recognition
title_sort few shot incremental learning with context aware spatial enhancement for image recognition
topic Few-shot incremental learning
image recognition
context-aware spatial enhancement
url https://ieeexplore.ieee.org/document/11036738/
work_keys_str_mv AT hengwu fewshotincrementallearningwithcontextawarespatialenhancementforimagerecognition
AT zeyang fewshotincrementallearningwithcontextawarespatialenhancementforimagerecognition
AT zijunzheng fewshotincrementallearningwithcontextawarespatialenhancementforimagerecognition
AT haiyangwang fewshotincrementallearningwithcontextawarespatialenhancementforimagerecognition
AT wansongwang fewshotincrementallearningwithcontextawarespatialenhancementforimagerecognition