Remote Sensing Scene Classification with Masked Image Modeling (MIM)
Not on the map. Why?
Abstract (arXiv)
Remote sensing scene classification has been extensively studied for its critical roles in geological survey, oil exploration, traffic management, earthquake prediction, wildfire monitoring, and intelligence monitoring. In the past, the Machine Learning (ML) methods for performing the task mainly used the backbones pretrained in the manner of supervised learning (SL). As Masked Image Modeling (MIM), a self-supervised learning (SSL) technique, has been shown as a better way for learning visual feature representation, it presents a new opportunity for improving ML performance on the scene classification task. This research aims to explore the potential of MIM pretrained backbones on four well-known classification datasets: Merced, AID, NWPU-RESISC45, and Optimal-31. Compared to the published benchmarks, we show that the MIM pretrained Vision Transformer (ViTs) backbones outperform other alternatives (up to 18% on top 1 accuracy) and that the MIM technique can learn better feature representation than the supervised learning counterparts (up to 5% on top 1 accuracy). Moreover, we show that the general-purpose MIM-pretrained ViTs can achieve competitive performance as the specially designed yet complicated Transformer for Remote Sensing (TRS) framework. Our experiment results also provide a performance baseline for future studies.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 59 of its 86 references that are in the map. First posted on arXiv 28 Feb 2023.
Builds on
59 of 86 references.
References · 59
- 2017Attention Is All You Needcited by 59,932
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
- 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
- 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
- 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
- 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,875
- 2021BEiT: BERT Pre-Training of Image Transformerscited by 2,360
- 2022VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Trainingcited by 1,305
- 2021SimMIM: A Simple Framework for Masked Image Modelingcited by 1,147
- 2021iBOT: Image BERT Pre-Training with Online Tokenizercited by 891
38 more
- 2017Remote Sensing Image Scene Classification: Benchmark and State of the Artcited by 791
- 2021Masked Feature Prediction for Self-Supervised Visual Pre-Trainingcited by 573
- 2022Masked Autoencoders As Spatiotemporal Learnerscited by 428
- 2016AID: A Benchmark Dataset for Performance Evaluation of Aerial Scene Classificationcited by 343
- 2022MCVD: Masked Conditional Video Diffusion for Prediction, Generation, and Interpolationcited by 307
- 2022Masked Siamese Networks for Label-Efficient Learningcited by 301
- 2022BEiT v2: Masked Image Modeling with Vector-Quantized Visual Tokenizerscited by 298
- 2022Context Autoencoder for Self-Supervised Representation Learningcited by 297
- 2022MultiMAE: Multi-modal Multi-task Masked Autoencoderscited by 287
- 2021BEVT: BERT Pretraining of Video Transformerscited by 169
- 2022Contrastive Masked Autoencoders are Stronger Vision Learnerscited by 133
- 2022What to Hide from Your Students: Attention-Guided Masked Image Modelingcited by 117
- 2021MST: Masked Self-Supervised Transformer for Visual Representationcited by 114
- 2022SemMAE: Semantic-Guided Masking for Learning Masked Autoencoderscited by 113
- 2022MaskViT: Masked Visual Pre-Training for Video Predictioncited by 112
- 2022ConvMAE: Masked Convolution Meets Masked Autoencoderscited by 105
- 2022OmniMAE: Single Model Masked Pretraining on Images and Videoscited by 99
- 2022MVP: Multimodality-guided Visual Pre-trainingcited by 92
- 2022How Mask Matters: Towards Theoretical Understandings of Masked Autoencoderscited by 86
- 2022Adversarial Masking for Self-Supervised Learningcited by 78
- 2022Masked Frequency Modeling for Self-Supervised Visual Pre-Trainingcited by 70
- 2022MILAN: Masked Image Pretraining on Language Assisted Representationcited by 69
- 2022MaskTune: Mitigating Spurious Correlations by Forcing to Explorecited by 66
- 2022Bootstrapped Masked Autoencoders for Vision BERT Pretrainingcited by 60
- 2022Uniform Masking: Enabling MAE Pre-training for Pyramid-based Vision Transformers with Localitycited by 57
- 2021VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learningcited by 56
- 2022SdAE: Self-distillated Masked Autoencodercited by 52
- 2019Deep Learning for Classification of Hyperspectral Data: A Comparative Reviewcited by 51
- 2022Green Hierarchical Vision Transformer for Masked Image Modelingcited by 50
- 2022Exploring Target Representations for Masked Autoencoderscited by 46
- 2022AdaMAE: Adaptive Masking for Efficient Spatiotemporal Learning with Masked Autoencoderscited by 41
- 2022Masked Autoencoders Enable Efficient Knowledge Distillerscited by 36
- 2022A Unified View of Masked Image Modelingcited by 36
- 2022Stare at What You See: Masked Image Modeling without Reconstructioncited by 30
- 2022MAR: Masked Autoencoders for Efficient Action Recognitioncited by 29
- 2022Efficient Self-supervised Vision Pretraining with Local Masked Reconstructioncited by 27
- 2022Exploring Long-Sequence Masked Autoencoderscited by 22
- 2022SupMAE: Supervised Masked Autoencoders Are Efficient Vision Learnerscited by 21
50 most cited of 59 All 59 (JSON)
Built on by
Citing papers per year: 2025: 2.
2 papers.