Data Fusion for Combining Information for Disparate Data Sources for Maritime Remote Sensing

Higgins, Erik; Sobien, Daniel; Freeman, Laura J.; Pitt, Jonathan S.

doi:10.2514/6.2021-0915

Cited by 4 publications

(6 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The SAR images are generated using end-to-end physics-based modeling and simulation tools developed to model complex phenomena on the ocean surface. The simulations are described by Higgins et al [17] but are summarized here: computational fluid dynamics (CFD) simulations model the evolution of a surface ship wake in a section of ocean; then, a second hydrodynamic simulation redistributes surfactants according to the surface currents and turbulence within the ship wake. More details about the hydrodynamic simulations that produced these SAR images can be found in [17], including the specific ship parameters, wind, ship speed, and headings.…”

Section: Ship Wake Sar Imagery Datasetmentioning

confidence: 99%

“…The simulations are described by Higgins et al [17] but are summarized here: computational fluid dynamics (CFD) simulations model the evolution of a surface ship wake in a section of ocean; then, a second hydrodynamic simulation redistributes surfactants according to the surface currents and turbulence within the ship wake. More details about the hydrodynamic simulations that produced these SAR images can be found in [17], including the specific ship parameters, wind, ship speed, and headings. Altering these parameters will impact the images generated and provide a larger data set, but this is currently beyond the scope of the work presented here.…”

Section: Ship Wake Sar Imagery Datasetmentioning

confidence: 99%

“…These radar cross section data are then used to construct the simulated SAR images. Variables including swell height and SAR platform parameters (see Table 1 from Higgins et al [17]), such as look angle and polarization, are systematically varied between the hydrodynamic and electromagnetic simulations to create realistic variation in the SAR images. The advantage of simulation is the ability to control these variables, which is especially helpful for studying the effects of data augmentation before applying it to real-world data, where there are more unknowns.…”

Section: Ship Wake Sar Imagery Datasetmentioning

confidence: 99%

“…Table 1. Run matrix for SAR image generation from [17]. The look angle is the azimuth angle between the radar direction and the ship direction, and the inclination angle is the angle between the vertical axis and the direction from the radar to the ship wake [18].…”

Section: Ship Wake Sar Imagery Datasetmentioning

confidence: 99%

“…In previous work [17], we developed three convolutional neural network (CNN) models for classifying the presence of a persistent wake in SAR imagery. Previous work has shown the effectiveness of CNN models for SAR imagery [7][8][9][10][11][12].…”

Section: Deep Learning Models For Ship Wake Detectionmentioning

confidence: 99%

See 4 more Smart Citations

Improving Deep Learning for Maritime Remote Sensing through Data Augmentation and Latent Space

Sobien

Higgins

Krometis

et al. 2022

MAKE

Self Cite

View full text Add to dashboard Cite

Training deep learning models requires having the right data for the problem and understanding both your data and the models’ performance on that data. Training deep learning models is difficult when data are limited, so in this paper, we seek to answer the following question: how can we train a deep learning model to increase its performance on a targeted area with limited data? We do this by applying rotation data augmentations to a simulated synthetic aperture radar (SAR) image dataset. We use the Uniform Manifold Approximation and Projection (UMAP) dimensionality reduction technique to understand the effects of augmentations on the data in latent space. Using this latent space representation, we can understand the data and choose specific training samples aimed at boosting model performance in targeted under-performing regions without the need to increase training set sizes. Results show that using latent space to choose training data significantly improves model performance in some cases; however, there are other cases where no improvements are made. We show that linking patterns in latent space is a possible predictor of model performance, but results require some experimentation and domain knowledge to determine the best options.

show abstract

Section: Ship Wake Sar Imagery Datasetmentioning

confidence: 99%

Section: Ship Wake Sar Imagery Datasetmentioning

confidence: 99%

Section: Ship Wake Sar Imagery Datasetmentioning

confidence: 99%

Section: Ship Wake Sar Imagery Datasetmentioning

confidence: 99%

Section: Deep Learning Models For Ship Wake Detectionmentioning

confidence: 99%

See 3 more Smart Citations

Improving Deep Learning for Maritime Remote Sensing through Data Augmentation and Latent Space

Sobien

Higgins

Krometis

et al. 2022

MAKE

Self Cite

View full text Add to dashboard Cite

show abstract

Development and Prospect of Satellite Remote Sensing Technology for Ocean Dynamic Environment

Zhang,

Jia

et al. 2023

Journal of Spacecraft and Rockets

View full text Add to dashboard Cite

Uncertainty Quantification in Data Fusion Classifier for Ship-Wake Detection

Costa,

Sobien,

Garg

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

Using deep learning model predictions requires not only understanding the model’s confidence but also its uncertainty, so we know when to trust the prediction or require support from a human. In this study, we used Monte Carlo dropout (MCDO) to characterize the uncertainty of deep learning image classification algorithms, including feature fusion models, on simulated synthetic aperture radar (SAR) images of persistent ship wakes. Comparing to a baseline, we used the distribution of predictions from dropout with simple mean value ensembling and the Kolmogorov—Smirnov (KS) test to classify in-domain and out-of-domain (OOD) test samples, created by rotating images to angles not present in the training data. Our objective was to improve the classification robustness and identify OOD images during the test time. The mean value ensembling did not improve the performance over the baseline, in that there was a –1.05% difference in the Matthews correlation coefficient (MCC) from the baseline model averaged across all SAR bands. The KS test, by contrast, saw an improvement of +12.5% difference in MCC and was able to identify the majority of OOD samples. Leveraging the full distribution of predictions improved the classification robustness and allowed labeling test images as OOD. The feature fusion models, however, did not improve the performance over the single SAR-band models, demonstrating that it is best to rely on the highest quality data source available (in our case, C-band).

show abstract

Data Fusion for Combining Information for Disparate Data Sources for Maritime Remote Sensing

Cited by 4 publications

References 10 publications

Improving Deep Learning for Maritime Remote Sensing through Data Augmentation and Latent Space

Improving Deep Learning for Maritime Remote Sensing through Data Augmentation and Latent Space

Development and Prospect of Satellite Remote Sensing Technology for Ocean Dynamic Environment

Uncertainty Quantification in Data Fusion Classifier for Ship-Wake Detection

Contact Info

Product

Resources

About