Loss Constraints Research Articles

Multilayer DS-MAC with game theory optimization

PurposeEnergy is the major concern in wireless sensor networks (WSNs) for most of the applications. There exist many factors for higher energy consumption in WSNs. The purpose of this work is to increase the coverage area maintaining the minimum possible nodes or sensors.Design/methodology/approachThis paper has proposed multilayer (ML) nodes deployment with distributed MAC (DS-MAC) in which nodes listen time is controlled based on communication of neighbors. Game theory optimization helps in addressing path loss constraints while selecting path toward base stations (BS).FindingsThe simulation is carried out using NS-2.35, and it shows better performance in ML DS-MAC compared to random topology in DS-MAC with same number of BS. The proposed method improves performance of network in terms of energy consumption, network lifetime and better throughput.Research limitations/implicationsEnergy consumption is the major problem in WSNs and for which there exist many reasons, and many approaches are being proposed by researchers based on application in which WSN is used. Node mobility, topology, multitier and ML deployment and path loss constraints are some of the concerns in WSNs.Practical implicationsGame theory is used in different situations like countries whose army race, business firms that are competing, animals generally fighting for prey, political parties competing for vote, penalty kicks for the players in football and so on.Social implicationsWSNs find applications in surveillance, monitoring, inspections for wild life, sea life, underground pipes and so on.Originality/valueGame theory optimization helps in addressing path loss constraints while selecting path toward BS.

Residual Refinement Network with Attribute Guidance for Precise Saliency Detection

As an important topic in the multimedia and computer vision fields, salient object detection has been researched for years. Recently, state-of-the-art performance has been witnessed with the aid of the fully convolutional networks (FCNs) and the various pyramid-like encoder-decoder frameworks. Starting from a common encoder-decoder architecture, we enhance a residual refinement network with feature purification for better saliency estimation. To this end, we improve the global knowledge streams with intermediate supervisions for global saliency estimation and design a specific feature subtraction module for residual learning, respectively. On the basis of the strengthened network, we also introduce an attribute encoding sub-network (AENet) with a grid aggregation block (GAB) to guide the final saliency predictor to obtain more accurate saliency maps. Furthermore, the network is trained with a novel constraint loss besides the traditional cross-entropy loss to yield the finer results. Extensive experiments on five public benchmarks show our method achieves better or comparable performance compared with previous state-of-the-art methods.

A Deep Non-negative Matrix Factorization Model for Big Data Representation Learning.

Nowadays, deep representations have been attracting much attention owing to the great performance in various tasks. However, the interpretability of deep representations poses a vast challenge on real-world applications. To alleviate the challenge, a deep matrix factorization method with non-negative constraints is proposed to learn deep part-based representations of interpretability for big data in this paper. Specifically, a deep architecture with a supervisor network suppressing noise in data and a student network learning deep representations of interpretability is designed, which is an end-to-end framework for pattern mining. Furthermore, to train the deep matrix factorization architecture, an interpretability loss is defined, including a symmetric loss, an apposition loss, and a non-negative constraint loss, which can ensure the knowledge transfer from the supervisor network to the student network, enhancing the robustness of deep representations. Finally, extensive experimental results on two benchmark datasets demonstrate the superiority of the deep matrix factorization method.

Scene Coordinate Regression Network With Global Context-Guided Spatial Feature Transformation for Visual Relocalization

Among visual relocalization from a single RGB image, the scene coordinate regression (SCoRe) based on convolutional neural network (CNN) becomes prevailing, however, it is insufficient to extract invariant features under different viewpoints due to fixed geometric structures of CNN. In this letter, we propose a global context-guided spatial feature transformation (SFT) network to learn invariant feature representation for robustness against viewpoint changes. Specifically, global feature extracted from source feature map is regarded as a dynamic convolutional kernel, which is convolved with source feature map for the prediction of transformation parameters. The predicted parameters are used to transform features of multiple viewpoints to a canonical space with the constraint of maximum likelihood-derived loss, and thus viewpoint invariance is achieved. CoordConv is also employed to further improve the discrimination of features on texture-less or repetitive zones. The proposed SFT network can be easily incorporated into the general SCoRe network. To our best knowledge, features are first decoupled from viewpoints explicitly in SCoRe network by the spatial feature transformation network, which achieves a stable and accurate visual relocalization. The experimental results demonstrate the effectiveness of the proposed method in terms of accuracy and efficiency.

Cardiac MRI segmentation with focal loss constrained deep residual networks

Delineating anatomical structures for cardiac magnetic resonance imaging (CMRI) is crucial for various medical applications such as medical diagnoses, treatment, and pathological studies. CMRI segmentation, which aims to automatically and accurately segment the heart structures, is highly beneficial for cardiologists. However, it is non-trivial to perfectly segment the ventricles, especially for the heart apex slices, considering their small sizes compared to the input images. For example, the endocardium in the Sunnybrook dataset only occupies 4% of the entire image by average. During the training process, these target pixels, or other hard samples, are buried by the massive backgrounds that make the model mostly receive optimization signals from easy samples. In this paper, we propose a focal loss constrained residual network (FR-Net) to tackle the problem. In order to mitigate the fact that the gradients of the hard samples can be easily overwhelmed by the easy samples, we use a pixel-wise re-weighting strategy to balance the gradients. Furthermore, considering focal loss constraints for each pixel independently, we propose an alternative training fashion that trains the model with focal loss and dice loss alternatively. The segmentation model can not only benefit from the pixel-wise focal loss but also from the region-wise dice loss to comprehensively optimize the model. We conducted thorough experiments on the Sunnybrook dataset, CMRI dataset, right ventricle dataset, and ACDC dataset to verify the effectiveness of the proposed method.

A novel quantization method combined with knowledge distillation for deep neural networks

The massive parameters and intensive computations in neural networks always limit the deployment on embedded devices with poor storage and computing power. To solve this problem, a novel quantization algorithm combined with Knowledge Distillation (KD) is proposed to reduce the model size and speed up the inference of deep models. The proposed method consists of two phases, KD-Training and Quantization-Retraining. KD-Training attempts to train a compact student model with pre-quantized weights by the proposed pre-quantized constraint loss. In Quantization-Retraining, the pre-quantized weights are quantized to 2n and the first and last layers of the network are retained to make up for the accuracy loss caused by quantization. Experiments on the CIFAR-10 dataset show that the proposed method can obtain a low-precision (2-5bit) quantized student model with a compact structure, and the test accuracy even exceeds its full-precision(32bit) reference as the improvement of the generalization ability. It can get higher performance compared with that obtained by other quantization methods. Also, since the quantized weights are constrained in { ± 2n}, it is suitable for the acceleration of the network calculation in hardware.

English

BACKGROUND Ankle fractures are among the most common injuries encountered by orthopaedic surgeons. An ankle is considered unstable when the loss of normal constraints around the ankle permits the talus to move in a non-physiologic pattern. Under such circumstances, the dynamic joint surface contact area within the ankle is diminished, which predisposes to articular cartilage damage and premature degeneration. In today’s time, many modalities are available for surgical fixation of bimalleolar fractures. Such interventions restore anatomy and biomechanics of the ankle joint. Even though there are plethora of foreign studies on similar matter, deficiency of such analysis has been depicted in literature from developing countries, particularly from South-East Asian region. In this study, we wanted to assess various methods of internal fixation in ankle fracture & evaluate their clinical and radiological outcome post-operatively. METHODS 26 ankle fracture patients in the age group of 18 - 60 years were included in this study. Fractures were classified pre-operatively based on Lauge-Hansen classification. Patients were followed up at regular intervals of 1st, 2nd, 3rd, 6th & 12 month after surgery and assessed by Baird and Jackson scoring system based on subjective, objective and radiographic criteria. Complications like infection, arthritis, stiffness & implant failure were assessed on regular follow-up. RESULTS Variety of methods (including k-wires, plates & screws) were used for surgical fixation. Majority of fractures resulted from road traffic accident (RTA). Patients were evaluated using Baird & Jackson scoring system during follow-up at 1st, 2nd , 3 rd, 6th & 12 month. All patients had poor scores for initial 3 months. Patients showed good to excellent result on final follow-up at 9 to 12 months after surgery. Complications were observed in 19 % cases with arthritis being most common. CONCLUSIONS Open reduction and internal fixation restores the articular congruity of the ankle joint. The operative result was satisfactory with good clinical outcome. KEYWORDS Ankle, Fracture, Anatomy, Modalities

Verb Second in Old Venetian

This study aims to add to a rich scholarship on the presence of a verb second constraint in old (Italo-)Romance that has been argued to cause V-to-C raising of both the finite verb and one or more constituents, provided we understand this constraint to be lax in these varieties (cf. Ledgeway 2007, 2008). In particular, it analyses a late 14th century old Venetian text, corroborating the existence of this constraint within the variety. Within this study, it will be shown that the syntax of 14th century Venetian is one in which we find the gradual loss of the V2 constraint. This will be shown through a range of phenomena present within the text, such as a loss of Informational Focus fronting, the weakening of subject pronouns, and the loss of scrambling to vP. All in all, this study aims to shed light on the syntax of 14th century Venetian specifically, adding to a growing scholarship on what the incipient loss of the V2 constraint looks like within respective old Romance varieties.

Constraints of Weight Loss as a Marker of Bariatric Surgery Success: An Exploratory Study.

PurposeThe aim of this exploratory study was to investigate whether the degree of weight loss properly reflects improvements in cardiometabolic health among patients who underwent Roux-en-Y gastric bypass.MethodsIn this ancillary analysis from a clinical trial, patients were clustered into tertiles according to the magnitude of the percentage weight loss (1st tertile: “higher weight loss”: −37.1 ± 5.8%; 2nd tertile: “moderate weight loss”: −29.7 ± 1.4%; 3rd tertile: “lower weight loss”: −24.2 ± 2.3%). Delta changes (9 months after surgery-baseline) in clustered cardiometabolic risk (i.e., blood pressure index, fasting glucose, high-density lipoprotein [HDL] and triglycerides [TG]), glycated hemoglobin (HbA1c), homeostasis model assessment (HOMA-IR), and C-reactive protein (CRP) were calculated.ResultsA total of 42 patients who had complete bodyweight data (age = 40 ± 8 year; BMI = 47.8 ± 7.1 kg/m2) were included. Surgery led to substantial weight loss (−37.9 ± 11.3 kg, P < 0,001), and clinically significant improvements in blood pressure index (−17.7 ± 8.2 mmHg, P < 0.001), fasting glucose (−36.6 ± 52.5 mg/dL, P < 0.001), HDL (9.4 ± 7.1 mg/dL, P < 0.001), TG (−35.8 ± 44.1 mg/dL P < 0,001), HbA1c (−1.2 ± 1.6%, P < 0.001), HOMA-IR (−4.7 ± 3.9 mg/dL, P < 0.001) and CRP (−8.5 ± 6.7 μg/mL P < 0.001). Comparisons across tertiles revealed no differences for cardiometabolic risk score, fasting glucose, HbAc1, HOMA-IR, blood pressure index, CRP, HDL, and TG (P > 0.05 for all). Individual variable analysis confirmed cardiometabolic improvements across the spectrum on weight-loss. There were no associations between weight loss and any dependent variable.ConclusionWeight loss following bariatric surgery does not correlate with improvements in cardiovascular risk factors. These findings suggest that weight loss alone may be insufficient to assess the cardiometabolic success of bariatric surgery, and the search for alternate proxies that better predict surgery success are needed.

Deformable adversarial registration network with multiple loss constraints.

Deformable medical image registration has the necessary value of theoretical research and clinical application. Traditional methods cannot meet clinical application standards in terms of registration accuracy and efficiency. This article proposes a deformable generate adversarial registration framework, which avoids the dependence on ground-truth deformation. The proposed residual registration network based on Nested U-Net has excellent feature extraction ability and robustness. Multiple constraints that incorporate the potential information of anatomical segmentation extracted by the discriminator can help the model adapt to different modal registration tasks. Through interpatient X-ray chest registration, the deep-supervised training method, and the proposed loss constraint are proved to improve the model's performance and training stability. The experimental results show that our model, compared with state-of-the-art methods, provides a more accurate spatial alignment relationship between different patients' lung organs while ensuring the displacement field's authenticity. Finally, we explored the relationship between the accuracy and validity of the model.

PCLoss: Fashion Landmark Estimation with Position Constraint Loss

Fashion landmark estimation aims at locating functional key points of clothes, which has wide potential applications in electronic commerce. However, due to the occlusion and weak outline information, landmark estimation occurs outliers and duplicate detection problems. To alleviate these issues, we propose Position Constraint Loss (PCLoss) to constrain error landmark locations by utilizing the position relationship of landmarks. Specifically, PCLoss adds a regularization term for each landmark to regularize their relative positions, and it can be easily applied to both regression and heatmap based methods without extra computation during inference. Unlike existing approaches that propagate landmark information between feature layers by specific network structures, PCLoss introduces position relations of landmarks in the label space without modifying the network structure. In addition, we leverage the skeleton-like relation of clothing to further strengthen position constraints between landmarks. Extensive experimental results on DeepFashion, FLD and FashionAI demonstrate that our methods can effectively increase the performance of mainstream frameworks by a large margin. We also explore the effectiveness of PCLoss on human pose estimation task, and the experimental results on COCO 2017 prove the generality of our methods on other key point estimation tasks.

Grammaticalization and language contact in a discourse-pragmatic change in progress: The spread ofinnitin London English

AbstractThis variationist analysis investigates the development and spread ofinnitas an invariant tag in London English. The sociolinguistic distribution ofinnitin a socially stratified corpus of vernacular speech suggests that the form's emergence and spread were initiated and propelled system-internally through changes associated with grammaticalization. Frequency triggered phonetic reduction ofisn't ittoinnit; loss of syntactic-semantic usage constraints and growing functional versatility enabledinnitto seize the range of contexts and functions of grammatically-dependent tags (e.g.didn't you,weren't we), virtually ousting these from the system of negative-polarity interrogative tags. Examination of cross-linguistic data and comparisons with relevant pre- and non-contact varieties indicate multiple language contact and grammatical replication may have played an ancillary role. I flag some challenges of establishing contact effects in discourse-pragmatic change, and propose that the promotion ofinnitfor invariant use was governed by its low salience and social indexicality of localness. (Innit, question tags, (Multicultural) London English, grammaticalization, language contact, grammatical replication)*

Extreme Y chromosome polymorphism corresponds to five male reproductive morphs of a freshwater fish.

Loss of recombination between sex chromosomes often depletes Y chromosomes of functional content and genetic variation, which might limit their potential to generate adaptive diversity. Males of the freshwater fish Poecilia parae occur as one of five discrete morphs, all of which shoal together in natural populations where morph frequency has been stable for over 50 years. Each morph uses a different complex reproductive strategy and morphs differ dramatically in colour, body size and mating behaviour. Morph phenotype is passed perfectly from father to son, indicating there are five Y haplotypes segregating in the species, which encode the complex male morph characteristics. Here, we examine Y diversity in natural populations of P. parae. Using linked-read sequencing on multiple P. parae females and males of all five morphs, we find that the genetic architecture of the male morphs evolved on the Y chromosome after recombination suppression had occurred with the X. Comparing Y chromosomes between each of the morphs, we show that, although the Ys of the three minor morphs that differ in colour are highly similar, there are substantial amounts of unique genetic material and divergence between the Ys of the three major morphs that differ in reproductive strategy, body size and mating behaviour. Altogether, our results suggest that the Y chromosome is able to overcome the constraints of recombination loss to generate extreme diversity, resulting in five discrete Y chromosomes that control complex reproductive strategies.

ТЕОРЕТИЧЕСКИЕ ОСНОВЫ ЭКОНОМИЧЕСКОГО И ИНСТИТУЦИОНАЛЬНОГО МЕХАНИЗМОВ УПРАВЛЯЕМОГО ПРИРОДОПОЛЬЗОВАНИЯ

Natural resources are the material basis for economic and social development, but the public characteristics of natural resources force the resource market to show more failures, the lack of a number of economic policies related to natural resources has led to the fact that the development, use and reproduction of natural resources has led to the loss of effective institutional constraints , which led to the overdevelopment, use and consumption of atmospheric resources, water resources, forest resources, land resources, and fuel and energy resources. which leads to the depletion of some resources, people can get local limited resources only by buying resources elsewhere. Therefore, it is necessary to improve control over natural resources and measures of the economic mechanism.&#x0D; The author of this article proposes methods for managing natural resources based on the characteristics of various indicators of natural resources (water, atmosphere, forest, land, fuel energy) and demonstrates the mechanisms of using natural resources in the main resource countries of the world.

Investigation on Beremin parameters based on the toughness scaling model

The toughness scaling model (TSM for short) is one of the methods to correct the constraint loss of small specimens. The principle of this method is the fracture toughness conversion at equal Weibull stress. However, related research shows that this fracture toughness scaling model can only accurately estimate the characteristic toughness J0. The scaled toughness distribution of small specimens always deviates from that of the standard specimens. In response to this widespread deviation, this paper studies the factors that significantly affect Weibull stress, such as the crack tip radius and parameter calibration methods. This research aims to determine the requirements for accurate fracture toughness conversion and minimize the deviation under the existing conditions. After studying the data of three different specimens, it is found that the TSM based on equal σw/σu and Minami's calibration method can accurately convert the toughness distribution of small specimens. However, the calibrated Weibull modulus of the two specimens must be close enough. Since it is difficult to guarantee that the calibrated parameters of the Beremin model are constant, the requirements for this accurate toughness conversion are relatively harsh. However, appropriately relaxing the Weibull modulus requirements can make the toughness scaling model more comfortable to build. Although the deviation cannot be avoided, it is still significantly reduced compared to the original toughness scaling model. On the other hand, this paper found that as long as the scale J0 of small specimens is equal to the J0 of standard specimens, an accurate reference temperature estimation can be easily obtained.

EVeREst: Bitrate Adaptation for Cloud VR

Cloud Virtual Reality (VR) technology is expected to promote VR by providing a higher Quality of Experience (QoE) and energy efficiency at lower prices for the consumer. In cloud VR, the virtual environment is rendered on the remote server and transmitted to the headset as a video stream. To guarantee real-time experience, networks need to transfer huge amounts of data with much stricter delays than imposed by the state-of-the-art live video streaming applications. To reduce the burden imposed on the networks, cloud VR applications shall adequately react to the changing network conditions, including the wireless channel fluctuations and highly variable user activity. For that, they need to adjust the quality of the video stream adaptively. This paper studies video quality adaptation for cloud VR and improves the QoE for cloud VR users. It develops a distributed, i.e., with no assistance from the network, bitrate adaptation algorithm for cloud VR, called the Enhanced VR bitrate Estimator (EVeREst). The algorithm aims to optimize the average bitrate of cloud VR video flows subject to video frame delay and loss constraints. For that, the algorithm estimates both the current network load and the delay experienced by separate frames. It anticipates the changes in the users’ activity and limits the bitrate accordingly, which helps prevent excess interruptions of the playback. With simulations, the paper shows that the developed algorithm significantly improves the QoE for the end-users compared to the state-of-the-art adaptation algorithms developed for MPEG DASH live streaming, e.g., BOLA. Unlike these algorithms, the developed algorithm satisfies the frame loss requirements of multiple VR sessions and increases the network goodput by up to 10 times.

Graph Convolutional Networks for the Automated Production of Building Vector Maps From Aerial Images

Building footprint delineation from remote sensing imagery is a basic task in surveying and mapping and geographic information system (GIS), which benefits many engineering applications but requires an enormous amount of human delineation. In this study, we aim to find a way to replace human delineation with automated algorithms. For this purpose, we designed a novel pipeline for the automated production of building vector maps from aerial images, which consists of a bounding-box generation module, a graph convolutional network (GCN)-based polygon prediction module, and an empirical polygon regularization module. First, we introduce the bounding box generation module based on region-based object detection, which along with our overlap cropping strategy is used to generate a bounding box for each building instance. The generated bounding boxes have a horizontal rectangle version and a rotated version, which are used to initialize the next stage of our method. Second, we propose a GCN-based method, which is the core of this study, to conduct the initial accurate building polygon prediction by integrating multiresolution optimization and multilevel loss constraints. Our proposed method, to the authors’ best knowledge, is the first of its kind that introduces GCN to building extraction. The final step is to apply a regularization algorithm to translate the predicted polygons into structured and highly accurate building footprints. We validated the proposed method on two large aerial building data sets, WHU data set and Inria data set, where it was shown to significantly outperform other state-of-the-art methods more than 10%. Specifically, with the ground-truth rotated bounding boxes of buildings, our method is able to automatically delineate 91% of the buildings in the WHU data set and with the predicted rotated bounding the percent reached human-level delineation.

A Rent-Seeking Framework for Multipath TCP

Network utility maximization (NUM) for Multipath TCP (MPTCP) is a challenging task, since there is no well-defined utility function for MPTCP [6]. In this paper, we identify the conditions under which we can use Kelly's NUM mechanism, and explicitly compute the equilibrium. We obtain this equilibrium by using Tullock's rent-seeking framework from game theory to define a utility function for MPTCP. This approach allows us to design MPTCP algorithms with common delay and/or loss constraints at the subflow level. Furthermore, this utility function has diagonal strict concavity, which guarantees a globally unique (normalized) equilibrium.

Self-learning soft computing algorithms for prediction machines of estimating crowd density

Human needs motivate the improvement of computing paradigms, the emergence of soft computing is more effective in dealing with daily problems, while reducing costs, it also solves the problem of robustness and becomes easier to handle. Examples of this include collecting data on video data from smart sensors and can use that information to predict and monitor the behavior of crowd. Nowadays, with the development of cyber–physical systems and artificial intelligence, the traditional data collection and analysis system faces the risks of low transparency and high data security, making it difficult to obtain an accurate prediction result. Therefore, in this paper a novel prediction machine via self-learning generative adversarial network for soft computing application is proposed, which collects data through a series of high-precision IoT sensor devices and makes preliminary preprocessing, and further solves the crowd prediction problem based on deep learning algorithms and obtains a reliable and accurate prediction result by continuously optimizing internal parameters. The focus of this work is on the accuracy and preprocessing of data collection and crowd prediction algorithms. The prediction algorithm can be used to estimate and monitor the crowd flow in public places, and can prevent crowding, trampling and other traffic jams, such as stations, airports, large exhibitions, tourist attractions and other places. Therefore, the new prediction machine includes video capture, upload and display, data analysis and early warning operations in embedded devices, and automatically predicts crowd density. In terms of constructing the network, first, in order to obtain a clearer generated density map, the feature self-learning module is merged in the generator feature extraction stage. Secondly, in order to avoid the blur of the generated image, an adversarial loss is constructed between the generator and the discriminator, and finally to deal with multiple scales, two generator networks are constructed to adapt to large scale and small scale respectively in order to extract semantic information of different scales and use cross-scale consistency loss constraints to generate density maps.

Fault-Tolerance Performance Analysis of a Five-Phase Permanent-Magnet Linear Synchronous Machine

A fault-tolerance control strategy based on the constant magnetomotive force (MMF) is proposed for a five-phase permanent-magnet linear synchronous machine (FP-PMLSM) under one-phase fault condition in this article, which aims to keep steady operation during a short time. First, the mathematical model of FP-PMLSM having one-phase open-circuit fault is analyzed. Based on the analysis, the fault-tolerance strategy is derived under the constraint of constant copper loss. This strategy makes the fault-tolerance currents become unbalanced seriously and the maximum thrust force reduced. Then, the proposed method is introduced and the corresponding fault-tolerance currents are calculated. The fault-tolerance currents of both strategies are optimized in order to reduce the second harmonic component of the thrust force ripple. Finally, the investigation of the electromagnetic performance is carried out in detail. By comparison with the constant copper loss strategy, the proposed strategy is proved to output a steadier thrust force.