Comparison of CNNs and ViTs Based Hybrid Models Using Gradient Profile Loss for Classification of Oil Spills in SAR Images

Oil spillage over a sea or ocean’s surface is a threat to marine and coastal ecosystems. Spaceborne synthetic aperture radar (SAR) data has been used efficiently for the detection of oil spills due to its operational capability in all-day all-weather conditions. The problem is often modeled as a semantic segmentation task. The images need to be segmented into multiple regions of interest such as sea surface, oil spill, look-alikes, ships and land. Training of a classifier for this task is particularly challenging since there is an inherent class imbalance. In this work, we train a convolutional neural network (CNN) with multiple feature extractors for pixel-wise classification; and introduce to use a new loss function, namely ‘gradient profile’ (GP) loss, which is in fact the constituent of the more generic Spatial Profile loss proposed for image translation problems. For the purpose of training, testing and performance evaluation, we use a publicly available dataset with selected oil spill events verified by the European Maritime Safety Agency (EMSA). The results obtained show that the proposed CNN trained with a combination of GP, Jaccard and focal loss functions can detect oil spills with an intersection over union (IoU) value of 63.95%. The IoU value for sea surface, look-alikes, ships and land class is 96.00%, 60.87%, 74.61% and 96.80%, respectively. The mean intersection over union (mIoU) value for all the classes is 78.45%, which accounts for a 13% improvement over the state of the art for this dataset. Moreover, we provide extensive ablation on different Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) based hybrid models to demonstrate the effectiveness of adding GP loss as an additional loss function for training. Results show that GP loss significantly improves the mIoU and F1 scores for CNNs as well as ViTs based hybrid models. GP loss turns out to be a promising loss function in the context of deep learning with SAR images.

Super Resolution of Magnetic Resonance Images

In this work, novel denoising and super resolution (SR) approaches for magnetic resonance (MR) images are addressed, and are integrated in a unified framework, which do not require example low resolution (LR)/high resolution (HR)/cross-modality/noise-free images and prior information of noise–noise variance. The proposed method categorizes the patches as either smooth or textured and then denoises them by deploying different denoising strategies for efficient denoising. The denoising algorithm is integrated into the SR approach, which uses a gradient profile-based constraint in a sparse representation-based framework to improve the resolution of MR images with reduced smearing of image details. This constraint regularizes the estimation of HR images such that the estimated HR image has gradient profiles similar to the gradient profiles of the original HR image. For this, the gradient profile sharpness (GPS) values of an unknown HR image are estimated using an approximated piece-wise linear relation among GPS values of LR and upsampled LR images. The experiments are performed on three different publicly available datasets. The proposed SR approach outperforms the existing unsupervised SR approach addressed for real MR images that exploits low rank and total variation (LRTV) regularization, by an average peak signal to noise ratio (PSNR) of 0.73 dB and 0.38 dB for upsampling factors 2 and 3, respectively. For the super resolution of noisy real MR images (degraded with 2% noise), the proposed approach outperforms the LRTV approach by an average PSNR of 0.54 dB and 0.46 dB for upsampling factors 2 and 3, respectively. The qualitative analysis is shown for real MR images from healthy subjects and subjects with Alzheimer’s disease and structural deformity, i.e., cavernoma.

Gradient Profile Estimation Using Exponential Cubic Spline Smoothing in a Bayesian Framework

Attaining reliable gradient profiles is of utmost relevance for many physical systems. In many situations, the estimation of the gradient is inaccurate due to noise. It is common practice to first estimate the underlying system and then compute the gradient profile by taking the subsequent analytic derivative of the estimated system. The underlying system is often estimated by fitting or smoothing the data using other techniques. Taking the subsequent analytic derivative of an estimated function can be ill-posed. This becomes worse as the noise in the system increases. As a result, the uncertainty generated in the gradient estimate increases. In this paper, a theoretical framework for a method to estimate the gradient profile of discrete noisy data is presented. The method was developed within a Bayesian framework. Comprehensive numerical experiments were conducted on synthetic data at different levels of noise. The accuracy of the proposed method was quantified. Our findings suggest that the proposed gradient profile estimation method outperforms the state-of-the-art methods.

Numerical and experimental study of the vertical distribution of velocity and hydraulic gradient in the capillary fringe

<p>Capillary fringe plays an important role in the fate and transport of infiltrated solutes from agricultural lands. In this study, flow patterns and the vertical distribution of the velocity and hydraulic gradient inside the capillary fringe were investigated using FEFLOW calibrated by experimental data. An experimental box along with a real sample of capillary fringe from the study area (Sand and clay pit Brüggen, Germany) was used for the experiments. The dimension of the filled part of the box was 0.75 m long, 0.55 m high, and 0.150 m wide. To maintain a constant hydraulic gradient throughout the experiments the upstream and downstream groundwater levels were fixed to 7 cm and 3 cm, respectively. The horizontal velocity at different points inside the capillary fringe and the vadose zone was measured by injecting the fluorescent dye tracer (Uranin). At the end of the experiments, the soil samples are collected from different parts of the box for water content measurement. The results indicate that FEFLOW successfully estimates water content, overall flow pattern, and more importantly horizontal movement inside the capillary fringe. The streamlines are parallel to the groundwater table in the middle part.  Based on both experimental and numerical results, while there is a downward movement near the outflow, an upward movement was seen near the inflow. In previous studies, the velocity profile inside the capillary fringe was estimated using Darcy’s law, unsaturated hydraulic conductivity, and constant hydraulic gradient. The detailed comparison of measured water content and velocity with numerical modeling results showed that the constant hydraulic gradient assumption above the water table in previous studies is not valid. The vertical hydraulic gradient profile calculated by FEFLOW showed that the hydraulic gradient at the middle part of the box changes from 0.042 to 0.03. Moreover, the shape of the vertical hydraulic gradient profile is a function of the location in the box and soil type.</p><p><strong>Keywords: </strong>Solute transport, Unsaturated zone, Streamline, Pore velocity, Hydraulic conductivity, FEFLOW</p>

ELECTRONIC CONTROL OF LOCAL UNIFORMITY OF ILLUMINANCE FROM POWER LED

<div class="contentJustify">It has been suggested to assess non-uniformity of illuminance by the coordinate distribution of its gradient. By the example of mathematical model, the possibility to form a desired illuminance gradient profile from a multicomponent module by means of electronic control of the LED lighting parameters.</div>