Correct Boundaries Research Articles

Medical image segmentation of gastric adenocarcinoma based on dense connection of residuals.

Accurate segmentation of gastric cancer based on CT images of gastric adenocarcinoma is crucial for physicians to screen gastric diseases, clinical diagnosis, preoperative prediction, and postoperative evaluation plans. To address the issue of the inability of the segmentation algorithm to depict the correct boundaries due to unclear gastric contours in the lesion area and the visible irregular band-like dense shadow extending to the perigastric region, a 3D medical image segmentation model 3D UNet based on residual dense jumping method is proposed. In the method we proposed, Residual Dense Block, which is applied to the image super-resolution module to remove CT artifacts, and Residual Block in ResNet are further fused. The quality of CT images is improved by Residual Dense Skip Block, which removes banded dense shadows, preserves image details and edge information, captures features, and improves the segmentation performance of gastric adenocarcinoma. The Instance Normalization layer position is modified to select the best result. Different loss functions are also combined in order to obtain the best gastric adenocarcinoma segmentation performance. We tested the model on a hospital-provided gastric adenocarcinoma dataset. The experimental results show that our model outperforms the existing methods in CT gastric adenocarcinoma segmentation, in which the method combining the hybrid loss function of Dice and CE obtains an average dice score of 82.3%, which is improved by 5.3% and 3.8% compared to TransUNet and Hiformer, respectively, and improves the cross-merge rate to 70.8%, compared to nnFormer, nnUNet by 1% and 0.9%, respectively. The residual jump connection structure indeed improves segmentation performance. The proposed method has the potential to be used as a screen for gastric diseases and to assist physicians in diagnosis.

Accurate contour preservation for semantic segmentation by mitigating the impact of pseudo-boundaries

Accurately extracting the contours of ground objects has been an important research topic in the field of semantic segmentation of remote sensing imagery. However, existing efforts have primarily focused on refining the boundaries of predictive masks, with little consideration given to pseudo boundaries caused by abrupt changes in surface textures. Therefore, this paper addresses this challenge with the contour preservation network (CPNet), a novel semantic segmentation network that effectively mitigates pseudo-boundary effects and produces more precise contours. The key of CPNet is the boundary-guided feature alignment module (BGAM). This module employs supervised boundary guidance to adaptively transfer the model’s attention from salient areas to correct semantic boundaries. This adaptive attention transfer mechanism enables the model to suppress the impact of internal pseudo boundaries and refine contours. To further refine boundaries, a boundary point feature rectification module (ReBPM) is designed to rectify the classification of boundary points with neighbor features. Extensive experimental validations have demonstrated the effectiveness and flexibility of the proposed CPNet on ISPRS Potsdam and Vaihingen datasets. The results showed that our model outperforms other state-of-the-art methods in terms of boundary IoU, mean IoU, and mean F1-score, and it exhibits significantly superior contour preservation ability compared to other models, notably in the presence of pseudo-boundaries. The code is available at: https://github.com/angiecao/CPNet.

POST-IVUS: A perceptual organisation-aware selective transformer framework for intravascular ultrasound segmentation.

Intravascular ultrasound (IVUS) is recommended in guiding coronary intervention. The segmentation of coronary lumen and external elastic membrane (EEM) borders in IVUS images is a key step, but the manual process is time-consuming and error-prone, and suffers from inter-observer variability. In this paper, we propose a novel perceptual organisation-aware selective transformer framework that can achieve accurate and robust segmentation of the vessel walls in IVUS images. In this framework, temporal context-based feature encoders extract efficient motion features of vessels. Then, a perceptual organisation-aware selective transformer module is proposed to extract accurate boundary information, supervised by a dedicated boundary loss. The obtained EEM and lumen segmentation results will be fused in a temporal constraining and fusion module, to determine the most likely correct boundaries with robustness to morphology. Our proposed methods are extensively evaluated in non-selected IVUS sequences, including normal, bifurcated, and calcified vessels with shadow artifacts. The results show that the proposed methods outperform the state-of-the-art, with a Jaccard measure of 0.92 for lumen and 0.94 for EEM on the IVUS 2011 open challenge dataset. This work has been integrated into a software QCU-CMS2 to automatically segment IVUS images in a user-friendly environment.

Experience of Belgorod region in carrying out complex cadastral works

The purpose of this article is to study the complex cadastral works and arising problems. In the process of writing the article, the authors used the Federal laws, the Land Code, the Urban Planning Code, special scientific literature, as well as working and statistical materials of the Belgorod region (from complex cadastral works in the Krasnensky and Prokhorovsky districts). To date, a large number of real estate objects throughout the Russian Federation have not been included in the Unified State Register of Real Estate, and some land plots do not have correct boundaries and exact areas preventing the successful maintenance of the Register and reliable determination of cadastral value for tax purposes.

Sequential sampling in prospective observational studies

Many factors must be taken into account when designing an observational study. Unlike controlled studies, observational studies cannot mitigate the effects of confounding through randomization, and such factors should be incorporated into both the study analysis and the study design. Unfortunately, there is often little data available on most of these factors at the design stage, rendering it infeasible to reliably postulate the impact of these factors on the treatment effect estimate and precision of such an effect. In this work we demonstrate how failure to account for adjustment covariates in the design stage of observational studies utilizing group sequential designs has deleterious effects on the study’s observed power. We do this by constructing a hypothetical study of Alzheimer’s disease biomarkers on cognition, using data from the Alzheimer’s Disease Neuroimaging Initiative. We then outline a constrained boundaries procedure that uses data collected at interim analyses to better estimate variance and correct stopping boundaries to maintain power and type I error, while allowing for early study termination.

Effects of inlet flow non-uniformities on thermochemical structures and quasi-one-dimensional simulation of sooting counterflow diffusion flames

Counterflow flames are routinely used for investigating fundamental flame and fuel properties such as laminar flame speeds, autoignition temperature, extinction strain rate, and chemistries of soot formation. The primary merit of counterflow flame is that the essentially two-dimensional configuration can be mathematically treated as a one-dimensional problem with certain assumptions made; this dimensional reduction is much beneficial for computational costs, which are critical for the investigation of complex chemistries such as those of soot formation. In this work, we performed a comprehensive investigation on the performance of the 1D modeling by comparing the results with experimental measurements and the more rigorous 2D models. We focused on the effects of inlet flow uniformities, which are frequencies assumed in the 1D model but challenging to realize in experiments. Parametric studies on the effects of nozzle flow rates, nozzle separation distances, and curtain flow rates on inlet flow uniformities and the 1D modeling were performed. The results demonstrated the importance to specify actual velocity boundary conditions, either obtained from experiments or from two-dimensional modeling to the 1D model. An additional novel contribution of this work is a quantitative presentation of the fact that the presence of the curtain flow would exert a notable influence on the core counterflow by modifying the radial distribution of the nozzle exit velocity although the effects can be accounted for by using the correct velocity boundaries in the quasi-1D model. This work provides recommendation for various geometry and operational parameters of the counterflow flame to facilitate researchers to select proper burner configuration and flow conditions that are amiable for accurate 1D modeling.

DPAM: A domain parser for AlphaFold models.

The recent breakthroughs in structure prediction, where methods such as AlphaFold demonstrated near-atomic accuracy, herald a paradigm shift in structural biology. The 200 million high-accuracy models released in the AlphaFold Database are expected to guide protein science in the coming decades. Partitioning these AlphaFold models into domains and assigning them to an evolutionary hierarchy provide an efficient way to gain functional insights into proteins. However, classifying such a large number of predicted structures challenges the infrastructure of current structure classifications, including our Evolutionary Classification of protein Domains (ECOD). Better computational tools are urgently needed to parse and classify domains from AlphaFold models automatically. Here we present a Domain Parser for AlphaFold Models (DPAM) that can automatically recognize globular domains from these models based on inter-residue distances in 3D structures, predicted aligned errors, and ECOD domains found by sequence (HHsuite) and structural (Dali) similarity searches. Based on a benchmark of 18,759 AlphaFold models, we demonstrate that DPAM can recognize 98.8% of domains and assign correct boundaries for 87.5%, significantly outperforming structure-based domain parsers and homology-based domain assignment using ECOD domains found by HHsuite or Dali. Application of DPAM to the massive AlphaFold models will enable efficient classification of domains, providing evolutionary contexts and facilitating functional studies.

Reducing Flipping Errors in Deep Neural Networks

Deep neural networks (DNNs) have been widely applied in various domains in artificial intelligence including computer vision and natural language processing. A DNN is typically trained for many epochs and then a validation dataset is used to select the DNN in an epoch (we simply call this epoch ``the last epoch") as the final model for making predictions on unseen samples, while it usually cannot achieve a perfect accuracy on unseen samples. An interesting question is ``how many test (unseen) samples that a DNN misclassifies in the last epoch were ever correctly classified by the DNN before the last epoch?". In this paper, we empirically study this question and find on several benchmark datasets that the vast majority of the misclassified samples in the last epoch were ever classified correctly before the last epoch, which means that the predictions for these samples were flipped from ``correct" to ``wrong". Motivated by this observation, we propose to restrict the behavior changes of a DNN on the correctly-classified samples so that the correct local boundaries can be maintained and the flipping error on unseen samples can be largely reduced. Extensive experiments on different benchmark datasets with different modern network architectures demonstrate that the proposed flipping error reduction (FER) approach can substantially improve the generalization, the robustness, and the transferability of DNNs without introducing any additional network parameters or inference cost, only with a negligible training overhead.

Simulation of statistical mean and variance of normally distributed random values, transformed by nonlinear functions $\sqrt{|X|}$ and $\sqrt{X}$

This paper presents theoretical studies of formation regularities for the statistical mean and variance of normally distributed random values with the unlimited argument values subjected to nonlinear transformations of functions $\sqrt{|X|}$ and $\sqrt{X}$. It is shown that for nonlinear square root transformation of a normally distributed random variable, the integrals of higher order mean $n>1$ satisfy the inequality $\overline{(y-\overline{Y})^n}\neq 0$. On the basis of the theoretical research, the correct boundaries $m,\sigma \to \infty$ of error transfer formulas are suggested.

Coupling of the meshless finite volume particle method and the finite element method for fluid–structure interaction with thin elastic structures

A novel fluid–structure interaction (FSI) scheme is developed in this article, with the meshless finite volume particle method (FVPM) for fluid dynamics and the finite element (FE) solver FEBio for solid mechanics. Like smoothed particle hydrodynamics (SPH), FVPM is based on compactly supported particle kernels, but kernels are interpreted as a partitioning of volume, leading to a scheme with properties of both SPH and classical finite volume methods. Particle methods are advantageous for fluid modelling in FSI of highly deformable structures, as the particles can conform to the dynamic geometry of the fluid–structure interface. In FVPM, in particular, particles are rigorously truncated by boundaries, and the novel FSI scheme incorporates a sharp interface between the fluid and solid. A new method is described which avoids the possibility of unphysical communication between fluid particles separated by a thin structure. The method proposed here ensures that particles interact with the correct neighbour particles and boundaries, regardless of the particle overlap with the thin structure, allowing for particle size that can be chosen independently of structure thickness. The new method is validated for a hydrostatic water column on an elastic plate, a dam-break with elastic gate, a dam-break with a downstream elastic wall, and a 2-D model of a heart valve leaflet. In the latter two cases, the structure thickness is less than the particle diameter by a factor of up to 16. The results agree well with experimental data from the literature for fluid velocity field and solid deformation. Where quantitative data is available, the numerical results converge with decreasing particle size.

Census Mapping in India and Role of GIS: A Look Ahead at Census 2021

The 2021 Census of India for over 1.3 billion population deploying 3 million enumerators, has significant evidence value for 71 countries where census is scheduled during 2021. Census mapping plays a major role in accurate, complete and timely census. It delineates the exact and correct boundaries of all the administrative units. The Indian census has been using Geographic Information System (GIS) technologies over the last three censuses. In this study, we focus on the applications and methodologies being adopted for the census mapping in Census 2021 in India which is going to be the first digital Census of India. Five mobile apps have been developed for data collection and for map-related work. The 2021 Indian census utilises the latest census mapping techniques, namely standardisation of GIS spatial database design, geo-referencing of administrative units and latest mobile mapping application (Arc GIS Quick Capture) for field operations and built-up area digitisation work. We also discuss the various challenges and their solutions for census mapping in India, most prominently a high quality, updated, comprehensive and geo-referenced address registry for accurate data collection and mapping, and the use of geo-referenced high-resolution satellite images at village level for covering the gaps in rural boundary maps.

Detecting Multi Word Terms in patents the same way as entities

In English patent document information retrieval, Multi Word Terms (MWTs) are an important factor in determining how relevant a patent document is for a particular search query. Detecting the correct boundaries for these MWTs is no trivial task and often complicated by the special writing style of the patent domain. In this paper we describe a method for detecting MWTs in patent sentences based on a method for detecting technical entities using deep learning. On our annotated dataset of 22 patents, our method achieved an average precision of 0.75, an average recall of 0.74 and an average F1 score of 0.74. Further, we argue for the use of domain specific word embedding resources and suggest that our model mostly learns whether individual words should be included in MWTs or not.

Significant of multi-level pre-processing steps and its proper sequence in SegCaps skin lesion segmentation of dermoscopic images

In computer-assisted diagnosis of dermoscopy images, skin lesion segmentation is a critical but difficult process. Many segmentation methods fail to extract correct lesion boundaries due to the lack of pre-processing steps or a very few number of pre-processing steps that aren't in the right order. This paper demonstrates the significant of Multi-level pre-processing steps and its proper sequence before SegCaps segmentation of Lesion images. In the proposed work author have planned four pre-processing steps before segmentation of skin lesion images in the sequence of first is a 2nd Order grey edge algorithm is used to achieve color consistency;. Second process consists of two sub-processes in the hair removal procedure. To get started, we use the Prewitt filter, which is adaptable to the hair pixel level. The following procedure is the Repairing Operation, which is carried out using the anisotropic Fast Marching Method. Third process is a noise filtering with the Improved Switching Bilateral Filter (ISBF), which uses the Sorted Quadrant Median Vector (SQMV) to determine best median value and the last process is The Directed Edge Enhancer algorithm is used to perform edge enhancement. The proposed model is tested Over MatlabR2017b, using the HAM10000 dataset. The proposed method's output is assessed by comparing segmentation effects with and without pre-processing, as well as visual appearances of pre-processed images from input images. From the results it has been concludes that when pre-processing steps are used in the correct order and proper sequence, segmentation accuracy is very high, as compared to segmentation without pre-processing steps.

Automatic Spectral Method of Mesh Segmentation Based on Fiedler Residual

In this paper, we propose a fully automatic mesh segmentation method, which divides meshes into sub-meshes recursively through spectral analysis. A common problem in the spectral analysis of geometric processing is how to choose the specific eigenvectors and the number of these vectors for analysis and processing. This method tackles this problem with only one eigenvector, i.e. Fiedler vector. In addition, using only one eigenvector drastically reduces the cost of computing. Different from the Fiedler vector commonly used in the bipartition of graphs and meshes, this method finds multiple parts in only one iteration, vastly reducing the number of iterations and thus the time of operation, because each iteration produces as many correct boundaries as possible, instead of only one. We have tested this method on many 3D models, the results of which suggest the proposed method performs better than many advanced methods of recent years.

Real-Time Shadow Detection From Live Outdoor Videos for Augmented Reality.

Simulating shadow interactions between real and virtual objects is important for augmented reality (AR), in which accurately and efficiently detecting real shadows from live videos is a crucial step. Most of the existing methods are capable of processing only scenes captured under a fixed viewpoint. In contrast, this article proposes a new framework for shadow detection in live outdoor videos captured under moving viewpoints. The framework splits each frame into a tracked region, which is the region tracked from the previous video frame through optical flow analysis, and an emerging region, which is newly introduced into the scene due to the moving viewpoint. The framework subsequently extracts features based on the intensity profiles surrounding the boundaries of candidate shadow regions. These features are then utilized to both correct erroneous shadow boundaries for the tracked region and to detect shadow boundaries for the emerging region by a Bayesian learning module. To remove spurious shadows, spatial layout constraints are further considered for emerging regions. The experimental results demonstrate that the proposed framework outperforms the state-of-the-art shadow tracking and detection algorithms on a variety of challenging cases in real time, including shadows on backgrounds with complex textures, nonplanar shadows, fast-moving shadows with changing typologies, and shadows cast by nonrigid objects. The quantitative experiments show that our method outperforms the best existing method, achieving a 33.3% increase in the average Fmeasure on a self-collected database. Coupled with an image-based shadow-casting method, the proposed framework generates realistic shadow interaction results. This capability will be particularly beneficial for supporting AR applications.

An MMT-Based Hierarchical Transmission Module for 4K/120fps Temporally Scalable Video

High frame rate (HFR) video is attracting strong interest since it is considered as a next step toward providing Ultra-High Definition video service. For instance, the Association of Radio Industries and Businesses (ARIB) standard, the latest broadcasting standard in Japan, defines a 120 fps broadcasting format. The standard stipulates temporally scalable coding and hierarchical transmission by MPEG Media Transport (MMT), in which the base layer and the enhancement layer are transmitted over different paths for flexible distribution. We have developed the first ever MMT transmitter/receiver module for 4K/120fps temporally scalable video. The module is equipped with a newly proposed encapsulation method of temporally scalable bitstreams with correct boundaries. It is also designed to be tolerant to severe network constraints, including packet loss, arrival timing offset, and delay jitter. We conducted a hierarchical transmission experiment for 4K/120fps temporally scalable video. The experiment demonstrated that the MMT module was successfully fabricated and capable of dealing with severe network constraints. Consequently, the module has excellent potential as a means to support HFR video distribution in various network situations.

EBIT: Weakly-supervised image translation with edge and boundary enhancement

Abstract In this paper, targeting image translation between the thermal and visible domains, we propose a novel framework to enhance the edge and boundary features during translation. We tackle the unsupervised training task where sample image pairs from two domains are randomly chosen so the image content does not match, but we also take advantage of paired feature maps during the feature disentanglement process. This can be considered to be weakly-supervisedtraining. First, since thermal images usually have vague edges, we propose to apply a Canny operator to strengthen the edge features of the thermal images. Next, to define the correct object boundaries, we extract the boundary features from silhouette masks which are paired with the visible domain images. Then we disentangle the features of both domains into a domain-shared latent space and a domain-exclusive latent space. In the domain-shared latent space, the edge and boundary features extracted from the thermal and visible domains respectively act as domain-shared information which is used to render the edges and define the boundaries of the translated images. In the domain-exclusive latent space, domain-exclusive information such as colour is used to render the colour of objects of the translated images. In addition, we propose a pixel-wise adversarial loss rather than more traditional ones. The experimental results show that the proposed method has the ability to render realistic edge and colour features within the correct object boundaries and outperforms several state-of-the-art methods.

LAKE ICE MONITORING WITH WEBCAMS AND CROWD-SOURCED IMAGES

Abstract. Lake ice is a strong climate indicator and has been recognised as part of the Essential Climate Variables (ECV) by the Global Climate Observing System (GCOS). The dynamics of freezing and thawing, and possible shifts of freezing patterns over time, can help in understanding the local and global climate systems. One way to acquire the spatio-temporal information about lake ice formation, independent of clouds, is to analyse webcam images. This paper intends to move towards a universal model for monitoring lake ice with freely available webcam data. We demonstrate good performance, including the ability to generalise across different winters and lakes, with a state-of-the-art Convolutional Neural Network (CNN) model for semantic image segmentation, Deeplab v3+. Moreover, we design a variant of that model, termed Deep-U-Lab, which predicts sharper, more correct segmentation boundaries. We have tested the model’s ability to generalise with data from multiple camera views and two different winters. On average, it achieves Intersection-over-Union (IoU) values of ≈71% across different cameras and ≈69% across different winters, greatly outperforming prior work. Going even further, we show that the model even achieves 60% IoU on arbitrary images scraped from photo-sharing websites. As part of the work, we introduce a new benchmark dataset of webcam images, Photi-LakeIce, from multiple cameras and two different winters, along with pixel-wise ground truth annotations.

One-vs-One classification for deep neural networks

For performing multi-class classification, deep neural networks almost always employ a One-vs-All (OvA) classification scheme with as many output units as there are classes in a dataset. The problem of this approach is that each output unit requires a complex decision boundary to separate examples from one class from all other examples. In this paper, we propose a novel One-vs-One (OvO) classification scheme for deep neural networks that trains each output unit to distinguish between a specific pair of classes. This method increases the number of output units compared to the One-vs-All classification scheme but makes learning correct decision boundaries much easier. In addition to changing the neural network architecture, we changed the loss function, created a code matrix to transform the one-hot encoding to a new label encoding, and changed the method for classifying examples. To analyze the advantages of the proposed method, we compared the One-vs-One and One-vs-All classification methods on three plant recognition datasets (including a novel dataset that we created) and a dataset with images of different monkey species using two deep architectures. The two deep convolutional neural network (CNN) architectures, Inception-V3 and ResNet-50, are trained from scratch or pre-trained weights. The results show that the One-vs-One classification method outperforms the One-vs-All method on all four datasets when training the CNNs from scratch. However, when using the two classification schemes for fine-tuning pre-trained CNNs, the One-vs-All method leads to the best performances, which is presumably because the CNNs had been pre-trained using the One-vs-All scheme.

Total variation diffusion and its application in shape decomposition

One challenge in shape decomposition is to capture correct boundaries between different parts and get piecewise constant results. Based on the good edge-preserving and sparsity properties of total variation regularization, this paper introduces a novel diffusion model by minimizing weighted total-variation energy with Dirichlet boundary constraints. By the total variation diffusion model, we propose an edge-preserving shape decomposition optimization model, which can be solved effectively by augmented Lagrangian method with each subproblem having closed form solution. A number of experiments display that our method can produce segmentation results with piecewise constant parts and feature-preserving boundaries for both meshes and 3D point clouds, especially for shapes with sharp features. In addition, for mesh segmentation, our results compare favorably to those obtained by several existing techniques when evaluated on the Princeton Segmentation Benchmark. Furthermore, the quantitative errors show that the algorithm is robust numerically and the computational costs are reasonable.