Evaluation of Split-Brain Autoencoders for High-Resolution Remote Sensing Scene Classification

By Vladan Stojnić, Vladimir Risojevic. 2018

12 citations recorded by the source corpus.

Self-supervised methods are interesting for remote sensing because there are not many human labeled datasets available, but there is practically unlimited amount of data that can be used for self-supervised learning. In this paper we analyze the use of split-brain autoencoders in the context of remote sensing image classification. Weinvestigate the importance of training set size, choice of color space and size of the model to the classification accuracy. We show that even with small amount of unlabeled training images, if we finetune the weights learned by the autoencoder, we can achieve almost state of the art results of 89.27% on AID dataset.

Subfields: Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Ji Qi
2. Vladimir Risojevic
3. Haifeng Li
4. Qing Zhu
5. Mingning Guo
6. Vladan Stojnić
7. Chao Tao

In Media Technology

1. Ji Qi
2. Vladimir Risojevic
3. Haifeng Li
4. Qing Zhu
5. Mingning Guo
6. Vladan Stojnić
7. Chao Tao

Citations in the corpus

• Self-Supervised Remote Sensing Feature Learning: Learning Paradigms, Challenges, and Future Works (2022, methodology): “Empirical experiments [138] show that the learned features using the laboratory space are more beneficial for downstream RSI scene classification tasks than using RGB space.”
• Self-Supervised Learning of Remote Sensing Scene Representations Using Contrastive Multiview Coding (2021, background): “In [31, 30] the authors analyzed the possibilities of using split-brain autoencoder [44] in the analysis of aerial images.”