Multiple Branches Research Articles

Parity-Time Symmetric Holographic Principle.

Originating from the Hamiltonian of a single qubit system, the phenomenon of the avoided level crossing is ubiquitous in multiple branches of physics, including the Landau-Zener transition in atomic, molecular, and optical physics, the band structure of condensed matter physics and the dispersion relation of relativistic quantum physics. We revisit this fundamental phenomenon in the simple example of a spinless relativistic quantum particle traveling in (1+1)-dimensional space-time and establish its relation to a spin-1/2 system evolving under a PT-symmetric Hamiltonian. This relation allows us to simulate 1-dimensional eigenvalue problems with a single qubit. Generalizing this relation to the eigenenergy problem of a bulk system with N spatial dimensions reveals that its eigenvalue problem can be mapped onto the time evolution of the edge state with (N-1) spatial dimensions governed by a non-Hermitian Hamiltonian. In other words, the bulk eigenenergy state is encoded in the edge state as a hologram, which can be decoded by the propagation of the edge state in the temporal dimension. We argue that the evolution will be PT-symmetric as long as the bulk system admits parity symmetry. Our work finds the application of PT-symmetric and non-Hermitian physics in quantum simulation and provides insights into the fundamental symmetries.

Fingerprinting Industrial IoT devices based on multi-branch neural network

Industrial Internet-of-Things systems suffer from a vast and vulnerable attack surface, raising widespread concerns about shielding IIoT devices from malicious attacks and reducing cyber risks. Device identification is the prerequisite to safeguard IIoT systems. We leverage the observation that IIoT network protocol implementations vary due to different hardware architectures/configurations and design tasks of IIoT devices, which cause the difference in their network traffic payloads. Specifically, we develop a novel neural network to learn the semantic/syntax features among multiple IIoT packets. The neural network has multiple branches, each of which consists of convolution layers, attention modules, and highway units for learning the classification model of IIoT devices. To validate the precision and recall of our neural network in IIoT devices fingerprinting, we have implemented a prototype of the proposed IIoT device identification system. Our results show that our approach achieves 95.8% precision and 95.4% recall, significantly outperforming other classification models.

Distinct spin and orbital dynamics in Sr2RuO4

The unconventional superconductor Sr2RuO4 has long served as a benchmark for theories of correlated-electron materials. The determination of the superconducting pairing mechanism requires detailed experimental information on collective bosonic excitations as potential mediators of Cooper pairing. We have used Ru L3-edge resonant inelastic x-ray scattering to obtain comprehensive maps of the electronic excitations of Sr2RuO4 over the entire Brillouin zone. We observe multiple branches of dispersive spin and orbital excitations associated with distinctly different energy scales. The spin and orbital dynamical response functions calculated within the dynamical mean-field theory are in excellent agreement with the experimental data. Our results highlight the Hund metal nature of Sr2RuO4 and provide key information for the understanding of its unconventional superconductivity.

Mechanical characteristics and reservoir stimulation mechanisms of the Gulong shale oil reservoirs, the northern Songliao Basin

Shale oil of the Qingshankou Formation of the Gulong Sag, the northern Songliao Basin, represents the first attempt at large-scale development of pure-shale-type shale oil in China. By integrating the multi-scale refined reservoir characterization with macro-micro-scale mechanical testing, it is clarified that the Gulong shale is characterized by high clay mineral content, high rock plasticity, highly-developed bedding, and prominent mechanical anisotropy. A three-dimensional (3D) fracture propagation model of hydraulic fracturing was built for the Gulong shale, which fully captures the hydraulic fracture distribution pattern affected by the high bedding density, in-situ stress, and fracturing treatment parameters. Our research showed that due to influences of bedding, hydraulic fracturing in the Gulong shale forms a complex fracture morphology featuring the main fracture with multiple perpendicular branches that have different lengths (like the outdoor directional TV antenna); however, the vertical propagation of fractures is inhibited, and the fracture height is commonly less than 10 m. The limited stimulated reservoir volume (SRV) is the main problem facing the fracturing stimulation of the Gulong shale oil. Bedding density has vital effects on fracture morphology, so case-specific fracturing designs shall be developed for shale intervals with different bedding development degrees. For reservoirs with well-developed bedding, it is suggested to properly increase the perforation cluster spacing and raise the volume and proportions of viscous fluids of the pad, so as to effectively promote vertical fracture propagation and improve reservoir stimulation performance. This study integrates multi-scale fine reservoir characterization and macro-micro-scale mechanical testing, as well as the construction and numerical simulation of hydraulic fracturing models for high-density layered shale reservoirs, providing a new approach and methodological framework for the fracturing research of high-density layered shale reservoirs.

Semisupervised Cross-Scale Graph Prototypical Network for Hyperspectral Image Classification.

In practice, the acquirement of labeled samples for hyperspectral image (HSI) is time-consuming and labor-intensive. It frequently induces the trouble of model overfitting and performance degradation for the supervised methodologies in HSI classification (HSIC). Fortunately, semisupervised learning can alleviate this deficiency, and graph convolutional network (GCN) is one of the most effective semisupervised approaches, which propagates the node information from each other in a transductive manner. In this study, we propose a cross-scale graph prototypical network (X-GPN) to achieve semisupervised high-quality HSIC. Specifically, considering the multiscale appearance of the land covers in the same remotely captured scene, we involve the neighborhoods of different scales to construct the adjacency matrices and simultaneously design a multibranch framework to investigate the abundant spectral-spatial features through graph convolutions. Furthermore, to exploit the complementary information between different scales, we simply employ the standard 1-D convolution to excavate the dependence of the intranode and concatenate the output with the features generated from other scales. Intuitively, different branches for various samples should have different importance to predict their categories. Thus, we develop a self-branch attentional addition (SBAA) module to adaptively highlight the most critical features produced by multiple branches. In addition, different from previous GCN for HSIC, we devise an innovative prototypical layer comprising a distance-based cross-entropy (DCE) loss function and a novel temporal entropy-based regularizer (TER), which can enhance the discrimination and representativeness of the node features and prototypes actively. Extensive experiments demonstrate that the proposed X-GPN is superior to the classic and state-of-the-art (SOTA) methods in terms of the classification performance.

Branch Aggregation Attention Network for Robotic Surgical Instrument Segmentation.

Surgical instrument segmentation is of great significance to robot-assisted surgery, but the noise caused by reflection, water mist, and motion blur during the surgery as well as the different forms of surgical instruments would greatly increase the difficulty of precise segmentation. A novel method called Branch Aggregation Attention network (BAANet) is proposed to address these challenges, which adopts a lightweight encoder and two designed modules, named Branch Balance Aggregation module (BBA) and Block Attention Fusion module (BAF), for efficient feature localization and denoising. By introducing the unique BBA module, features from multiple branches are balanced and optimized through a combination of addition and multiplication to complement strengths and effectively suppress noise. Furthermore, to fully integrate the contextual information and capture the region of interest, the BAF module is proposed in the decoder, which receives adjacent feature maps from the BBA module and localizes the surgical instruments from both global and local perspectives by utilizing a dual branch attention mechanism. According to the experimental results, the proposed method has the advantage of being lightweight while outperforming the second-best method by 4.03%, 1.53%, and 1.34% in mIoU scores on three challenging surgical instrument datasets, respectively, compared to the existing state-of-the-art methods. Code is available at https://github.com/SWT-1014/BAANet.

Learn by Oneself: Exploiting Weight-Sharing Potential in Knowledge Distillation Guided Ensemble Network

Recent CNNs (convolutional neural networks) have become more and more compact. The elegant structure design highly improves the performance of CNNs. With the development of knowledge distillation technique, the performance of CNNs gets further improved. However, existing knowledge distillation guided methods either rely on offline pretrained high-quality large teacher models or online heavy training burden. To solve the above problems, we propose a feature-sharing and weight-sharing based ensemble network (training framework) guided by knowledge distillation (EKD-FWSNet) to make baseline models stronger in terms of representation ability with less training computation and memory cost involved. Specifically, motivated by getting rid of the dependence of offline pretrained teacher model, we design an end-to-end online training scheme to optimize EKD-FWSNet. Motivated by decreasing the online training burden, we only introduce one auxiliary classmate branch to construct multiple forward branches, which will then be integrated as ensemble teacher to guide baseline model. Compared to previous online ensemble training frameworks, EKD-FWSNet can provide diverse output predictions without relying on increasing auxiliary classmate branches. Motivated by maximizing the optimization power of EKD-FWSNet, we exploit the representation potential of weight-sharing blocks and design efficient knowledge distillation mechanism in EKD-FWSNet. Extensive comparison experiments and visualization analysis on benchmark datasets (CIFAR-10/100, tiny-ImageNet, CUB-200 and ImageNet) show that self-learned EKD-FWSNet can boost the performance of baseline models by large margin, which has obvious superiority compared to previous related methods. Extensive analysis also proves the interpretability of EKD-FWSNet. Our code is available at https://github.com/cv516Buaa/EKD-FWSNet.

Abstract 111: Woven Endo Bridge‐Coil Embolization of a Middle Cerebral Artery Bifurcation Aneurysm with Secondary Lobule

Introduction 18‐36% of all intracranial aneurysms are middle cerebral artery(MCA) bifurcation aneurysms. While some are much higher risk than others, they have an overall low estimated annual rupture rate of 0.36%.(1) MCA bifurcation aneurysms have been historically challenging to treat endovascularly given they often have complex morphology and wide necks. The Woven EndoBridge (WEB) is an ellipsoid embolization device designed to provide intrasaccular flow disruption along the aneurysm neck, which has been proven to be an effective method in treating a wide spectrum of wide‐necked bifurcation aneurysms. Thromboembolism and recanalization are the main concerns in treatment with the WEB device, like any aneurysm treatment approach .(2),(3) Occasionally, irregularities and secondary lobules at the base of the primary aneurysm can have persistent filling despite treatment. Methods We present a case of a right MCA bifurcation aneurysm treated by a WEB device and coil embolization. Results 48‐year‐old female with a medical history significant only for active smoking who presented with sudden onset of the worst headache of her life followed by brief loss of consciousness. No neurological evaluation was sought initially, until 2 weeks later as she continued to experience a moderate‐intensity headache associated with nausea and vomiting with frequent falls. On exam, she was alert and oriented, moving all extremities with full strength. CT head revealed an evolving acute to subacute right MCA infarction with adjacent small volume subarachnoid hemorrhage. CTA head revealed a right MCA bifurcation aneurysm measuring 6 x 6 x 7 mm with irregularity of the wall. Multifocal severe stenosis was noted in the right distal right M1 and right M2 divisions. She underwent a diagnostic subtraction angiography which confirmed an irregular right MCA bifurcation aneurysm measuring 4.3 x 4.5 x 5.7 mm (Figure 1 ‐A,B). While it was believed that the subarachnoid hemorrhage was more likely related to the recent infarct, a decision was made to treat the aneurysm given its relatively large size and irregular shape. Embolization was performed with a WEB SL 6 x 4 mm device (Figure 1: C,D ‐blue arrow). Post‐deployment angiograph showed occlusion of the aneurysm dome but persistent filling of the secondary lobule near the base of the aneurysm. Given this residual lobule filling, a decision was made to continue additional embolization with a hyper soft 3D 2 mm X 2 cm coil(Figure 1: C,D ‐ yellow arrow). Final angiography revealed complete occlusion of the aneurysm. Vasospasm of the distal M2 divisions and multiple M3 branches in the right MCA vascular territory was also noted. Therefore, she was treated with 20 mg of intra‐arterial Verapamil with improvement. She was started on Aspirin 325 mg daily. Daily transcranial dopplers were negative for vasospasm. She remained neurologically stable and was discharged home 5 days after admission ambulating independently. Conclusion With modern advances in endovascular devices, many MCA bifurcation aneurysms can be treated minimally invasively. However, there are some limitations to the WEB since every aneurysm is different and there are a finite number of sizes and shapes available for WEB. This case illustrates that adjunctive coil embolization can successfully occlude a residual secondary base lobule after WEB embolization.

Current Tracking and Circulating Current Suppression Super-Twisting Sliding Mode Control for Parallel Operation of EAST Fast Control Power Supply

To quickly output sufficient current for plasma excitation control, parallel operation of a structure of multiple branches is adopted in the Experimental Advanced Superconducting Tokamak (EAST) fast control power supply. During the process of parallel operation of multiple branches in engineering, a larger inductance current sharing reactor is used to suppress the circulating current for the branches, which reduces the dynamic response speed of the output current and increases economic costs. In order to achieve cost savings and improve the dynamic response speed of the output current, the parallel branch current model of the EAST fast control power supply is analyzed, and the current of each branch is reconstructed into two parts: the current flowing to the load end and the circulating current flowing to other branches. Without changing the circuit structure and increasing the additional complex communication system for each branch, observation of the current flowing to the load end from each branch is achieved. Based on the observed current, a super-twisting sliding mode controller (STSMC) is designed to suppress the circulating current flowing through branches. To realize fast output of the branch current and circulating current suppression for the branches, a new STSMC with a linear term and parameter adaptive structure is designed, speeding up the convergence rate of the whole control system. The linear term and designed parameter adaptive structure based on the sliding mode system status ensure fast convergence speed and excellent control performance of the system. Simulation and experiments show that the designed control method can achieve fast output current control for each branch and that the tracking performance of the total output current is good. While reducing the inductance of the current sharing reactor, the circulating current for the branches is effectively suppressed compared with traditional control methods. The proposed method has great significance in cost savings and performance improvement in engineering practice applications.

Machine-learning-assisted wearable PVA/Acrylic fluorescent layer-based triboelectric sensor for motion, gait and individual recognition

Flexible sensors with precise perception ability play an important role in environmental sensing. The organic fusion between the sensitive structure and sensitive material presents some challenges for the overall design and integration of the sensor, including designing multiple sensitive branch structures, simplifying system design, and sensing multiple stress modes. Here, based on a same-plane interdigital electrode design, we developed a triboelectric sensor and its array (TSA) that mimics the synaptic structure of a neural cell. The interdigital electrode structure can effectively simulate the synaptic structure of neural cells and achieve effective perception of multi-directional stress. The enhancement of the triboelectric effect is achieved through layer-by-layer self-assembly technology between PVA film and acrylic fluorescent layer. The interface effect between PVA film and acrylic fluorescent layer can effectively enhance the electrostatic effect. Further analysis with deep learning CNN can extract higher-level motion features. A single sensor can be used to identify which part of the body is producing the movement (98.89%). A TSA with different numbers of sensors can achieve high-precision recognition of gesture numbers (99.25%), foot shape, gait (99.23%), and individuals (99.75%). The sensor can be flexibly manufactured and integrated according to specific application scenarios. The long-term goal is to effectively help limb-injured patients with rehabilitation therapy and build intelligent home environment.

Outcomes of Castor Single-Branched Stent Graft for Reconstruction of Multiple Supra-Aortic Branches in Aortic Arch Disease.

To report the outcomes of a combination of Castor single-branched stent grafts with other techniques for the reconstruction of multiple supra-aortic branches in aortic arch disease. Between December 2019 and December 2021, 20 patients with aortic arch disease underwent thoracic endovascular aortic repair (TEVAR) at our institution using a Castor single-branched stent graft combined with the fenestration, chimney, or bypass techniques. Thoracic endovascular aortic repair is indicated for complicated or acute type B aortic dissection (TBAD), nonruptured aneurysms with a maximum aneurysm diameter >5.5 cm or showing rapidly expanded, ruptured, or threatened aneurysms, and penetrating aortic ulcers (PAUs) with a maximal aortic diameter >5.5 cm or with PAUs >10 mm deep or >20 mm in diameter. Preoperative, intraoperative, and postoperative clinical data were recorded. The median age of the patients was 56 (range=52-69 years) years, and 19 patients were men. Seven patients underwent the Castor single-branched stent graft and left common carotid artery (LCCA) chimney technique, 8 patients underwent the Castor single-branched stent graft and fenestration technique, and 5 patients underwent the Castor single-branched stent graft and bypass technique. The technical success rate was 100%. Major adverse events included 2 endoleaks, 1 spinal cord ischemia, and 1 early-stage retrograde type A aortic dissection. No cerebral stroke-related complications were observed. The mortality rate was 10% (2/20 patients). One patient with thoracic aortic aneurysm (TAA) died because of a sudden decrease in oxygen saturation and blood pressure after surgery. Relatives declined autopsy, and the cause of death was not determined. Another patient died of a retrograde type A dissection after surgery. The median follow-up period was 6 months (range=3.5-12 months). During follow-up, 1 patient with type I endoleak underwent thoracotomy again after a year. The remaining patients recovered well. The combination of a Castor single-branched stent graft with fenestration, chimney, or bypass techniques may be an effective treatment for preserving multiple supra-aortic branches in aortic arch disease. This study introduced three methods of reconstruction of multiple supra-aortic branches using a Castor single-branched stent graft (Castor single-branched stent graft combined with fenestration, chimney, or bypass technique) and analysed their advantages and shortcomings to provide experience for the future treatment of aortic arch diseases.

Numerical studies on the effects of melting heat transfer over quadratic radiative-magnetic hybrid nanofluid flow toward a porous shrinkable surface influenced by second-order slip velocity: Stability analysis and entropy generation

ABSTRACT The exclusive behavior of hybrid nanofluid flow and heat transmission properties through a porous medium toward a porous shrinking surface along with the external magnetic field, melting heat transfer, and second-order slip is investigated in this study. Moreover, the effects of a porous medium and quadratic thermal radiation are also incorporated. Molybdenum disulfide () and silicon dioxide () are hybridized as a very dilute homogenous mixture in the base fluid water (). Nonlinear partial differential equations with specified boundary constraints are transformed into a set of nonlinear ordinary differential equations using similarities transformation before being solved numerically with the help of default solver by bvp4c technique in MATLAB programming. This work is distinctive because it presents a novel hybrid nanofluid mathematical model with multiple (lower and upper) solution branches for the shrinking case within the defined ranges of the physical parameters. The temporal stability analysis shows that only one solution is stable as time evolves. The results are shown as velocity and temperature curves, skin friction coefficient, and local Nusselt number in the form of tabulated data and graphical structures. The heat for quadratic thermal radiation and melting heat transfer parameters contribute to the progress of thermal enhancement. Furthermore, by combining silicon oxide with a volume percentage of molybdenum disulfide, the flow is retarded, and heat transfer is enhanced. Entropy formation is considerably enhanced by quadratic thermal radiation and nanoparticle solid volume fraction parameters. In engineering, hybrid nanofluid has many uses, including in cosmetics, fabrics, paper plastics, automobiles, soaps, food packaging, food colorants, homes, cancer treatment, pharmaceuticals, ceramics, and paints.

A novel distributing–collecting on-line insulation monitoring system and line selection technology for the DC supply system

At present, the power supply system of 5G base stations is a micro smart grid, it generally uses 240 V DC power supply with multiple branches, and leakage accidents will threaten personal and property safety, so it is vital to identify the fault line accurately and remove the faults rapidly. In this paper, the leakage phenomenon of transmission lines in the HVDC power supply system of a 5G communication base station is studied. To address the issue of multi-branch line leakage diagnosis and line selection in the 240 V DC system, a new distributed DC insulation monitoring and fault line selection technology and system are proposed. The distributed DC insulation monitoring line selection technology is used to collect the leakage current of each branch by setting a unique single-core DC leakage transformer and process the signal primarily. In addition, a high-resistance bridge switchgear is added to the bus bar. According to different leakage currents caused by the imbalance of the bridge before and after the switch, the insulation resistance of electrodes to the ground is calculated accurately, which solves the problem that the current technology cannot judge the fault of the positive and negative electrodes to the ground at the same time. Through both simulation and prototype experiments, the feasibility of this technology and system device in line insulation monitoring and line selection of the HVDC communication base station power supply system is verified.

A Mathematical Connection Between Single-Elimination Sports Tournaments and Evolutionary Trees

Summary How many ways are there to arrange the sequence of games in a single-elimination sports tournament? We consider the connection between this enumeration problem and the enumeration of “labeled histories,” or sequences of asynchronous branching events, in mathematical phylogenetics. The possibility of playing multiple games simultaneously in different arenas suggests an extension of the enumeration of labeled histories to scenarios in which multiple branching events occur simultaneously. We provide a recursive result enumerating game sequences and labeled histories in which simultaneity is allowed. For a March Madness basketball tournament of 68 labeled teams, the number of possible sequences of games is ∼ 1.91 × 10 78 if arbitrarily many arenas are available, but only ∼ 3.60 × 10 68 if all games must be played sequentially in the same arena.

Scientific Knowledge Map Study of Therapeutic Landscapes and Community Open Spaces: Visual Analysis with CiteSpace

The pursuit of a healthy and comfortable living environment is a key developmental objective for human society. Therapeutic landscapes play a significant role in improving environmental conditions within these spaces. However, current research suggests that there is still much to be explored in this field, particularly in communal open spaces. Based on the Web of Science literature database and using the CiteSpace visualization tool, this study launched a literature review search on the existing research content on therapeutic landscapes in community spaces by means of publication volume analysis, spatial distribution, keyword analysis, co-citation clustering analysis, keyword timeline, and co-occurring word analysis. Based on the research results, the current status and direction of related research are outlined, and the research hotspots and future trends in this field are analyzed. Current research comprises multiple interdisciplinary branches, such as geography, public space, modern medicine, care, horticultural therapy, urban ecology, and more, with theoretical research, caregiving, spatial territories, and research methodology as the main research vectors. It is clear from this study that the current research on community therapeutic landscapes suffers from a lack of coordination between theoretical and practical development, and the related design practice activities are in a vulnerable stage of development. In terms of the population served, specialized research will be one of the directions of development, as there has been a gradual increase in the number of spatial research on the prevention and complementary treatment of various diseases for subdivided groups. At the same time, the research focus in this field has shifted from the physical health of users to their mental health, leading to a trend of public service development with the objective of social health.

ARiRTN: A Novel Learning-Based Estimation Model for Regressing Illumination.

In computational color constancy, regressing illumination is one of the most common approaches to manifesting the original color appearance of an object in a real-life scene. However, this approach struggles with the challenge of accuracy arising from label vagueness, which is caused by unknown light sources, different reflection characteristics of scene objects, and extrinsic factors such as various types of imaging sensors. This article introduces a novel learning-based estimation model, an aggregate residual-in-residual transformation network (ARiRTN) architecture, by combining the inception model with the residual network and embedding residual networks into a residual network. The proposed model has two parts: the feature-map group and the ARiRTN operator. In the ARiRTN operator, all splits perform transformations simultaneously, and the resulting outputs are concatenated into their respective cardinal groups. Moreover, the proposed architecture is designed to develop multiple homogeneous branches for high cardinality, and an increased size of a set of transformations, which extends the network in width and in length. As a result of experimenting with the four most popular datasets in the field, the proposed architecture makes a compelling case that complexity increases accuracy. In other words, the combination of the two complicated networks, residual and inception networks, helps reduce overfitting, gradient distortion, and vanishing problems, and thereby contributes to improving accuracy. Our experimental results demonstrate this model's outperformance over its most advanced counterparts in terms of accuracy, as well as the robustness of illuminant invariance and camera invariance.

Stability of non-Newtonian nanofluid movement with heat/mass transportation passed through a hydro magnetic elongating/contracting sheet: multiple branches solutions

Nanomaterials have found wide applications in many fields, leading to significant interest in the scientific world, in particular automobile thermal control, heat reservoirs, freezers, hybrid control machines, paper creation, cooling organisms, etc. The aim of the present study is to investigate the MHD non-Newtonian nanofluid and time-based stability analysis to verify the stable branch by computing the smallest eigenvalue across a slendering, extending, or shrinking sheet with thermal radiation and chemical reactions. The basic flow equations have been obtained in terms of PDEs, which are then converted to ODEs in dimensionless form via a suitable transformation. Based on the MATLAB software package bvp4c, the numerical solution has been obtained for the system of equations. A comparative study of the present and published work is impressive. The influence of evolving factors such as Prandtl number, Schmidt number, magnetic factor, heat generation/absorption, thermal, thermophoresis factor, chemical factor, second-grade fluid factor, and Brownian number on the velocities, energy, and concentration patterns is discussed through graphs. It is perceived that the temperature distribution enriches owing to the greater magnitude of the heat source. Furthermore, it is observed that a greater magnitude of radiation improves the temperature curves. It is also investigated from the present analysis that concentration and temperature profiles increase due to the growing values of the thermophoresis factor.

Seismic attenuation tomography of Eastern Europe from ambient seismic noise analysis

SUMMARY The Eastern-Europe region (EER), is a complex geotectonic area that captures part of the Alpine-Himalayan Orogen, the subduction of multiple NeoTethys Branches and part of the East European Craton. It is one of the most exciting geological areas in Europe due to a diversity of tectonic processes acting within it: extensional basin evolution, oceanic subduction, post-collisional volcanism, as well as active crustal deformation associated with the push of the Adria plate or the pull of the actively detaching Vrancea slab. This makes EER an excellent natural laboratory to study the behaviour of the lithosphere–asthenosphere system in a heterogeneous tectonic setting. To investigate the lateral heterogeneity and physical properties of the crust in the EER, we use ambient seismic noise data recorded by the vertical components of broad-band stations that have been operational between 1999 and 2020 in Eastern Europe and surrounding regions. We used this significant amount of data and the latest processing techniques of the ambient seismic noise field based on the continuous wavelet transform to compute cross-correlations between various station pairs, turning every available seismic station into a virtual source. The coda of the interstation cross-correlograms were used to determine coda quality factors (Qc) of Rayleigh waves in four different period ranges (3.0–5.0, 5.0–10.0, 10.0–20.0 and 20.0–30.0 s) and to invert them in the 2-D space, constructing the highest resolution attenuation tomography of the region. Our results reveal high attenuation features throughout the northeast Pannonian region, the Bohemian Massif, the East Carpathians and the Moesian Platform. Nevertheless, our findings do not emphasize a close correlation between the depth of sedimentary basins and attenuation features identified at longer periods. In addition, Qc variations are larger at short periods, indicating higher heterogeneity in the uppermost crust of Eastern Europe. Our findings demonstrate the higher efficiency of noise correlation approaches relative to earthquake data analyses investigating Qc at low frequencies.

HAMCap: A Weak-Supervised Hybrid Attention-Based Capsule Neural Network for Fine-Grained Climate Change Debate Analysis

Climate change (CC) has become a central global topic within the multiple branches of social disciplines. Natural Language Processing (NLP) plays a superior role since it has achieved marvelous accomplishments in various application scenarios. However, CC debates are ambiguous and complicated to interpret even for humans, especially when it comes to the aspect-oriented fine-grained level. Furthermore, the lack of large-scale effective labeled datasets is always a plight encountered in NLP. In this work, we propose a novel weak-supervised Hybrid Attention Masking Capsule Neural Network (HAMCap) for fine-grained CC debate analysis. Specifically, we use vectors with allocated different weights instead of scalars, and a hybrid attention mechanism is designed in order to better capture and represent information. By randomly masking with a Partial Context Mask (PCM) mechanism, we can better construct the internal relationship between the aspects and entities and easily obtain a large-scale generated dataset. Considering the uniqueness of linguistics, we propose a Reinforcement Learning-based Generator-Selector mechanism to automatically update and select data that are beneficial to model training. Empirical results indicate that our proposed ensemble model outperforms baselines on downstream tasks with a maximum of 50.08% on accuracy and 49.48% on F1 scores. Finally, we draw interpretable conclusions about the climate change debate, which is a widespread global concern.

Fiber Optic Sensor Based on Linear Sagnac Interferometer for Branch Localization

One of the challenges of distributed fiber optic sensors is to monitor objects that have multiple branches, such as pipelines or power grids. Existing sensors require multiple sensing systems to cover each branch, which increases the cost and complexity of the monitoring. We propose a fiber optic sensor based on linear Sagnac interferometer for branch localization for practical environments where branch paths exist. The system is based on a dual-wavelength single-core Sagnac interferometer, which can extend multiple branches in the sensing cable part. The phase difference signals caused by external disturbances are demodulated and go through different signal processing. By using time delay estimation and null frequency method, we can determine the exact location of the disturbance on the main or branch path. Moreover, we propose a new method to reduce interference with null frequencies of the spectrum by subtracting the phase difference signal at one wavelength from the other. We demonstrate the feasibility and performance of our proposed system through theoretical analysis and experimental results. Our proposed system offers a simple and effective solution for branch localization in distributed fiber optic sensing applications.