Modern approaches in deep learning for SAR ATR

Wilmanski, Michael; Kreucher, Chris; Lauer, J. L.

doi:10.1117/12.2220290

Cited by 43 publications

(29 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This section reviews some of the relevant studies in this area. Many authors applied CNN to SAR ATR and tested on MSTAR dataset, e.g., [43][44][45][46], etc.…”

Section: B Interpretation Of Sar Imagesmentioning

confidence: 99%

Deep Learning in Remote Sensing: A Comprehensive Review and List of Resources

Zhu

Tuia

Mou

et al. 2017

IEEE Geosci. Remote Sens. Mag.

2,232

1,204

View full text Add to dashboard Cite

This is the pre-acceptance version, to read the final version please go to IEEE Geoscience and Remote Sensing Magazine on IEEE XPlore.Standing at the paradigm shift towards data-intensive science, machine learning techniques are becoming increasingly important. In particular, as a major breakthrough in the field, deep learning has proven as an extremely powerful tool in many fields. Shall we embrace deep learning as the key to all? Or, should we resist a "black-box" solution? There are controversial opinions in the remote sensing community. In this article, we analyze the challenges of using deep learning for remote sensing data analysis, review the recent advances, and provide resources to make deep learning in remote sensing ridiculously simple to start with. More importantly, we advocate remote sensing scientists to bring their expertise into deep learning, and use it as an implicit general model to tackle unprecedented large-scale influential challenges, such as climate change and urbanization. Following this wave of success and thanks to the increased availability of data and computational resources, the use of deep learning in remote sensing is finally taking off in remote sensing as well. Remote sensing data bring some new challenges for deep learning, since satellite image analysis raises some unique questions that translate into challenging new scientific questions:• Remote sensing data are often multi-modal, e.g. from optical (multi-and hyperspectral) and synthetic aperture radar (SAR) sensors, where both the imaging geometries and the content are completely different. Data and information fusion uses these complementary

show abstract

“…This section reviews some of the relevant studies in this area. Many authors applied CNN to SAR ATR and tested on MSTAR dataset, e.g., [43][44][45][46], etc.…”

Section: B Interpretation Of Sar Imagesmentioning

confidence: 99%

Deep Learning in Remote Sensing: A Comprehensive Review and List of Resources

Zhu

Tuia

Mou

et al. 2017

IEEE Geosci. Remote Sens. Mag.

2,232

1,204

View full text Add to dashboard Cite

show abstract

“…Morgan et al [31] proposed an architecture of three convolutional layers, following a fully connected layer of Softmax as a classifier, increasing the accuracy to 92.3%. Moreover, Wilmanski et al [32] explored different learning algorithms of training CNNs, finding that the AdaDelta technique that can update the various learning rates of hyper-parameters outperformed the other techniques such as stochastic gradient descent (SGD) and AdaGrad. Recently, in [24], a five-layer all-convolutional network was proposed.…”

Section: Sar Target Recognition With Cnnsmentioning

confidence: 99%

Transfer Learning with Deep Convolutional Neural Network for SAR Target Classification with Limited Labeled Data

2017

View full text Add to dashboard Cite

Abstract:Tremendous progress has been made in object recognition with deep convolutional neural networks (CNNs), thanks to the availability of large-scale annotated dataset. With the ability of learning highly hierarchical image feature extractors, deep CNNs are also expected to solve the Synthetic Aperture Radar (SAR) target classification problems. However, the limited labeled SAR target data becomes a handicap to train a deep CNN. To solve this problem, we propose a transfer learning based method, making knowledge learned from sufficient unlabeled SAR scene images transferrable to labeled SAR target data. We design an assembled CNN architecture consisting of a classification pathway and a reconstruction pathway, together with a feedback bypass additionally. Instead of training a deep network with limited dataset from scratch, a large number of unlabeled SAR scene images are used to train the reconstruction pathway with stacked convolutional auto-encoders (SCAE) at first. Then, these pre-trained convolutional layers are reused to transfer knowledge to SAR target classification tasks, with feedback bypass introducing the reconstruction loss simultaneously. The experimental results demonstrate that transfer learning leads to a better performance in the case of scarce labeled training data and the additional feedback bypass with reconstruction loss helps to boost the capability of classification pathway.

show abstract

“…A CNN architecture modified from Wilmanski et al [3] is selected as a baseline to compare with the proposed method. The differences between the baseline and the original Wilmanski's network are a size of input image and kernel sizes of three convolutional layers.…”

Section: Baselinementioning

confidence: 99%

“…Several classification approaches have already been proposed in SAR ATR. Wilmanski et al applied convolutional neural network (CNN) to SAR classification problem and showed promising results of high accuracy [3]. The method achieved 97 % in accuracy while manual feature based classifiers showed approximately 70 %.…”

Section: Introductionmentioning

confidence: 99%

Rollable Latent Space for Azimuth Invariant Sar Target Recognition

Sagi

Toizumi

Senda

2018

IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium

View full text Add to dashboard Cite

This paper proposes rollable latent space (RLS) for an azimuth invariant synthetic aperture radar (SAR) target recognition. Scarce labeled data and limited viewing direction are critical issues in SAR target recognition.The RLS is a designed space in which rolling of latent features corresponds to 3D rotation of an object. Thus latent features of an arbitrary view can be inferred using those of different views. This characteristic further enables us to augment data from limited viewing in RLS. RLS-based classifiers with and without data augmentation and a conventional classifier trained with target front shots are evaluated over untrained target back shots. Results show that the RLS-based classifier with augmentation improves an accuracy by 30 % compared to the conventional classifier.

show abstract

Modern approaches in deep learning for SAR ATR

Cited by 43 publications

References 7 publications

Deep Learning in Remote Sensing: A Comprehensive Review and List of Resources

Deep Learning in Remote Sensing: A Comprehensive Review and List of Resources

Transfer Learning with Deep Convolutional Neural Network for SAR Target Classification with Limited Labeled Data

Rollable Latent Space for Azimuth Invariant Sar Target Recognition

Contact Info

Product

Resources

About