Deep Learning Earth Observation Classification Using ImageNet Pretrained Networks

By Dimitrios Marmanis, Mihai Datcu, Thomas Esch, Uwe Stilla. 2016

581 citations recorded by the source corpus.

Subfields: Artificial Intelligence, 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 Artificial Intelligence

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification (2017, background): “As shown in previous work [6], [15], [17], [19], deep CNNs have been shown to outperform non-deep learning approaches in land use and land cover image classification.”
Pre-Trained AlexNet Architecture with Pyramid Pooling and Supervision for High Spatial Resolution Remote Sensing Image Scene Classification (2017, methodology): “From Figure 19, it can be seen that the pre-trained AlexNet-SPP-SS model performs better over the training sample ratios of [10, 20, 30, 40, 50] than the pre-trained AlexNet, the pre-trained AlexNet-SPP, and the pre-trained AlexNet-SS models for the Google image dataset of SIRI-WHU.”
A Deeply Supervised Attention Metric-Based Network and an Open Aerial Image Dataset for Remote Sensing Change Detection (2021, background): “to automatically extract hierarchical multilevel features with rich spectral and spatial features from satellite images [20].”
A Transdisciplinary Review of Deep Learning Research and Its Relevance for Water Resources Scientists (2017, methodology): “For example, publicly available trained networks such as GoogLeNet (Szegedy et al., 2015) can be transferred for scene classification for satellite images (Hu et al., 2015; Marmanis et al., 2016; Nogueira et al., 2017).”
The color out of space: learning self-supervised representations for Earth Observation imagery (2020, background): “In these contexts, ﬁne-tuning pre-trained models often mitigate the lack of a large annotated dataset, yielding great performance in some cases [19], [20].”
Semantic segmentation of slums in satellite images using transfer learning on fully convolutional neural networks (2019, unknown): “…learning enables to further improve the learning task where backpropagation successfully re-weights labeled data from natural image datasets, e.g. ImageNet to solve new problems in remote sensing datasets (e.g. Maggiori et al., 2017; Marmanis et al., 2016; Nogueira et al., 2017; Kang et al., 2018).”
Remote Sensing Image Scene Classification With Self-Supervised Paradigm Under Limited Labeled Samples (2020, background): “Although many research works [1], [5], [6] have demonstrated that the features extracted from pretrained convolutional neural networks (CNNs) on ImageNet can generalize well to aerial image scene classification tasks, pretraining CNNs on ImageNet has two limitations.”
Remote Sensing Image Scene Classification Meets Deep Learning: Challenges, Methods, Benchmarks, and Opportunities (2020, unknown): “In [108], Marmanis et al.”
In-domain representation learning for remote sensing (2019, methodology): “This is a relatively small datasets that has been widely benchmarked for remote sensing scene classification task since 2010 and for which nearly perfect accuracy can be achieved with modern convolutional neural networks (Castelluccio et al., 2015; Marmanis et al., 2016; Nogueira et al., 2017).”
Learning and Transferring Deep Joint Spectral–Spatial Features for Hyperspectral Classification (2017, background): “Recently, transferring pretrained deep learning network has shown its potential in remote-sensed scene classification [41]–[43].”
Functional Map of the World (2017, background): “Recently, there have been several, mostly successful, attempts to leverage techniques that were founded on first-person imagery and apply them to remote sensing data [15, 21, 30].”
Research Progress on Few-Shot Learning for Remote Sensing Image Interpretation (2021, unknown): “For the scene classiﬁcation task in optical images, Marmanis et al. [87] designed a two-stage transfer learning framework to investigate the potential of using the large pretrained neural network for earth observation classiﬁcation.”
Deep Learning for Land Use and Land Cover Classification Based on Hyperspectral and Multispectral Earth Observation Data: A Review (2020, methodology): “References [141,142] use a classifier pre-trained on the ImageNet dataset to transfer knowledge into a land cover classification problem.”
Deep learning in remote sensing: a review (2017, background): “The deep arhitectures discussed in the “Convolutional Neural Networks” section have been applied to the scene classification problem of high-resolution satellite images and led to state-of-the-art performance [71], [74], [80]–[87].”