Unsupervised Few-Shot Continual Learning for Remote Sensing Image Scene Classification
Not on the map. Why?
Abstract (arXiv)
A continual learning (CL) model is desired for remote sensing image analysis because of varying camera parameters, spectral ranges, resolutions, etc. There exist some recent initiatives to develop CL techniques in this domain but they still depend on massive labelled samples which do not fully fit remote sensing applications because ground truths are often obtained via field-based surveys. This paper addresses this problem with a proposal of unsupervised flat-wide learning approach (UNISA) for unsupervised few-shot continual learning approaches of remote sensing image scene classifications which do not depend on any labelled samples for its model updates. UNISA is developed from the idea of prototype scattering and positive sampling for learning representations while the catastrophic forgetting problem is tackled with the flat-wide learning approach combined with a ball generator to address the data scarcity problem. Our numerical study with remote sensing image scene datasets and a hyperspectral dataset confirms the advantages of our solution. Source codes of UNISA are shared publicly in urlhttps://github.com/anwarmaxsum/UNISA to allow convenient future studies and reproductions of our numerical results.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 49 of its 81 references that are in the map. First posted on arXiv 4 Jun 2024.
Builds on
49 of 81 references.
References · 49
- 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
- 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
- 2020Exploring Simple Siamese Representation Learningcited by 2,853
- 2016Learning without Forgettingcited by 2,465
- 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,313
- 2017Gradient Episodic Memory for Continual Learningcited by 1,871
- 2018Unsupervised Feature Learning via Non-Parametric Instance-level Discriminationcited by 1,838
- 2017EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classificationcited by 1,488
- 2016Progressive Neural Networkscited by 1,332
- 2018Continual Lifelong Learning with Neural Networks: A Reviewcited by 1,313
- 2017Memory Aware Synapses: Learning what (not) to forgetcited by 1,094
- 2021Learning to Prompt for Continual Learningcited by 938
37 more
- 2020Dark Experience for General Continual Learning: a Strong, Simple Baselinecited by 895
- 2017Remote Sensing Image Scene Classification: Benchmark and State of the Artcited by 791
- 2017Lifelong Learning with Dynamically Expandable Networkscited by 689
- 2022DualPrompt: Complementary Prompting for Rehearsal-free Continual Learningcited by 635
- 2019Three scenarios for continual learningcited by 610
- 2018Progress & Compress: A scalable framework for continual learningcited by 516
- 2022CODA-Prompt: COntinual Decomposed Attention-based Prompting for Rehearsal-Free Continual Learningcited by 446
- 2016AID: A Benchmark Dataset for Performance Evaluation of Aerial Scene Classificationcited by 343
- 2019Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgettingcited by 302
- 2020Few-Shot Class-Incremental Learningcited by 249
- 2021Few-Shot Incremental Learning with Continually Evolved Classifierscited by 184
- 2019Continual Unsupervised Representation Learningcited by 147
- 2020Adversarial Continual Learningcited by 140
- 2021Overcoming Catastrophic Forgetting in Incremental Few-Shot Learning by Finding Flat Minimacited by 118
- 2023Neural Collapse Inspired Feature-Classifier Alignment for Few-Shot Class Incremental Learningcited by 112
- 2020CPR: Classifier-Projection Regularization for Continual Learningcited by 62
- 2021Few-Shot Lifelong Learningcited by 58
- 2021Learning Representation for Clustering via Prototype Scattering and Positive Samplingcited by 58
- 2023Introducing Language Guidance in Prompt-based Continual Learningcited by 49
- 2023Few-Shot Class-Incremental Learning via Training-Free Prototype Calibrationcited by 44
- 2022On the Soft-Subnetwork for Few-shot Class Incremental Learningcited by 37
- 2019Overcoming Catastrophic Forgetting by Neuron-level Plasticity Controlcited by 36
- 2024Convolutional Prompting meets Language Models for Continual Learningcited by 32
- 2019Unsupervised Progressive Learning and the STAM Architecturecited by 29
- 2022Single-source Domain Expansion Network for Cross-Scene Hyperspectral Image Classificationcited by 25
- 2024OrCo: Towards Better Generalization via Orthogonality and Contrast for Few-Shot Class-Incremental Learningcited by 25
- 2022Reinforced Continual Learning for Graphscited by 24
- 2023S3C: Self-Supervised Stochastic Classifiers for Few-Shot Class-Incremental Learningcited by 23
- 2022Language-aware Domain Generalization Network for Cross-Scene Hyperspectral Image Classificationcited by 20
- 2021Unsupervised Continual Learning via Self-Adaptive Deep Clustering Approachcited by 19
- 2021Unsupervised Continual Learning in Streaming Environmentscited by 19
- 2021Continual Learning via Inter-Task Synaptic Mappingcited by 14
- 2022Class-Incremental Learning via Knowledge Amalgamationcited by 9
- 2023Assessor-Guided Learning for Continual Environmentscited by 7
- 2022Scalable Adversarial Online Continual Learningcited by 6
- 2023Few-Shot Continual Learning via Flat-to-Wide Approachescited by 4
- 2024Few-Shot Class Incremental Learning via Robust Transformer Approachcited by 4
Built on by
Citing papers per year: 2025: 1.
1 paper.