# Training general representations for remote sensing using in-domain knowledge

arXiv: https://arxiv.org/abs/2010.00332 · HTML: https://mlsanity.com/paper/2010.00332/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 30 Sep 2020.

## Built on by (most cited first, 12 of 12)

- [SSL4EO-S12: A Large-Scale Multi-Modal, Multi-Temporal Dataset for Self-Supervised Learning in Earth Observation](https://mlsanity.com/paper/2211.07044/) (2022)
- [Revisiting pre-trained remote sensing model benchmarks: resizing and normalization matters](https://mlsanity.com/paper/2305.13456/) (2023)
- [Current Trends in Deep Learning for Earth Observation: An Open-source Benchmark Arena for Image Classification](https://mlsanity.com/paper/2207.07189/) (2022)
- [Semantic-aware Dense Representation Learning for Remote Sensing Image Change Detection](https://mlsanity.com/paper/2205.13769/) (2022)
- [Contrastive Multiview Coding with Electro-optics for SAR Semantic Segmentation](https://mlsanity.com/paper/2109.00120/) (2021)
- [Do we still need ImageNet pre-training in remote sensing scene classification?](https://mlsanity.com/paper/2111.03690/) (2021)
- [Semantic decoupled representation learning for remote sensing image change detection](https://mlsanity.com/paper/2201.05778/) (2022)
- [Self-Supervised In-Domain Representation Learning for Remote Sensing Image Scene Classification](https://mlsanity.com/paper/2302.01793/) (2023)
- [Learning with less: label-efficient land cover classification at very high spatial resolution using self-supervised deep learning](https://mlsanity.com/paper/2511.03004/) (2025)
- [Source data selection for out-of-domain generalization](https://mlsanity.com/paper/2202.02155/) (2022)
- [Supervised and Contrastive Self-Supervised In-Domain Representation Learning for Dense Prediction Problems in Remote Sensing](https://mlsanity.com/paper/2301.12541/) (2023)
- [Semantics-Aware Hierarchical Consensus Learning for Remote Sensing Image Classification](https://mlsanity.com/paper/2510.04916/) (2025)

## Abstract (arXiv)

Automatically finding good and general remote sensing representations allows to perform transfer learning on a wide range of applications - improving the accuracy and reducing the required number of training samples. This paper investigates development of generic remote sensing representations, and explores which characteristics are important for a dataset to be a good source for representation learning. For this analysis, five diverse remote sensing datasets are selected and used for both, disjoint upstream representation learning and downstream model training and evaluation. A common evaluation protocol is used to establish baselines for these datasets that achieve state-of-the-art performance. As the results indicate, especially with a low number of available training samples a significant performance enhancement can be observed when including additionally in-domain data in comparison to training models from scratch or fine-tuning only on ImageNet (up to 11% and 40%, respectively, at 100 training samples). All datasets and pretrained representation models are published online.