Two-dimensional Representation Research Articles

Analysis of the performance of a hybrid CPU/GPU 1D2D coupled model for real flood cases

Abstract Coupled 1D2D models emerged as an efficient solution for a two-dimensional (2D) representation of the floodplain combined with a fast one-dimensional (1D) schematization of the main channel. At the same time, high-performance computing (HPC) has appeared as an efficient tool for model acceleration. In this work, a previously validated 1D2D Central Processing Unit (CPU) model is combined with an HPC technique for fast and accurate flood simulation. Due to the speed of 1D schemes, a hybrid CPU/GPU model that runs the 1D main channel on CPU and accelerates the 2D floodplain with a Graphics Processing Unit (GPU) is presented. Since the data transfer between sub-domains and devices (CPU/GPU) may be the main potential drawback of this architecture, the test cases are selected to carry out a careful time analysis. The results reveal the speed-up dependency on the 2D mesh, the event to be solved and the 1D discretization of the main channel. Additionally, special attention must be paid to the time step size computation shared between sub-models. In spite of the use of a hybrid CPU/GPU implementation, high speed-ups are accomplished in some cases.

Proximal Over-Resection During Femoral Osteochondroplasty Negatively Affects the Distractive Stability of the Hip Joint

Objectives:Hip microinstability is an increasingly recognized source of hip pain and disability. Femoral osteochondroplasty is usually performed with direct visualization through the arthroscope, assisted with repeated fluoroscopic view for orientation and feedback. However, a two-dimensional representation of a three dimensional structure can be misleading compromising the precision of the planned osteochondroplasty. Sometimes the resection can extend proximally into the Femoral Head (FH) diminishing the surface area of the articular cartilage available to be in contact with the labrum, in order to create the suction effect (suction seal) that maintains the stability of the femoral head in the acetabulum. The purpose of the study is to evaluate the role of proximal over-resection of femoral osteochondroplasty in the rotational and distractive stability of the hip joint. We hypothesis that proximal over resection will result in decrease stability, specially at higher degrees of flexion and internal rotation where the contact between labrum and articular cartilage will be lost, and breaking the suction seal.Methods:Six hemi-pelvises were repeatedly tested in the following five conditions: (a) intact, (b) T-capsulotomy, (c) Osteochondroplasty to the level of the physeal scar, followed by a (d) 5mm and (e) 10mm proximal extension of the resection. The pelvis was secured to a metal plate and the distal portion of the femur was potted and attached to a multi-axial hip jig. (Fig.1) Specimens were axially distracted with a load from 0-150N followed by 5Nm of internal and external torque at 0 o, 15 o, 30 o, 60 o, 90o of flexion while the resultant displacement/rotation was recorded using a 3D motion tracking system. Repeated measures ANOVA was used with statistical significance set at p<0.05.Results:Proximal extension of the resection by 5mm and 10mm increased axial instability (decreased force required for hip distraction) at every angle of flexion tested, with the greatest increase observed at higher angles of flexion (60 deg and 90 deg), p<0.05. T-capsulotomy alone increased both internal and external rotation at all angles of flexion, p<0.05. (Fig. 2) Subsequent resection and extension of the resection did not significantly increase rotational instability compared to the capsulotomy state.Conclusion:Extending the osteochondroplasty proximally into the femoral head compromised the distractive stability of the hip joint. It is important to be precise when performing femoral osteochondroplasty to minimize proximal extension that can lead to iatrogenic instability of the hip joint and poor postoperative outcomes.Figure 1.Figure 2.

Eco-Efficient Value Creation of Residential Street Lighting Systems by Simultaneously Analysing the Value, the Costs and the Eco-Costs during the Design and Engineering Phase

In search of sustainable business models, product innovation must fulfil a double objective: the new product must have a higher (market) value, and at the same time a lower eco-burden. To achieve this objective, it is an imperative that the value, the total costs of ownership, and the eco-burden of a product are analysed at the beginning of the design process (idea generation and concept development). The design approach that supports such a design objective, is called Eco-efficient Value Creation (EVC). This approach is characterised by a two-dimensional representation: the eco-burden at the y-axis and the costs or the value at the x-axis. The value is either the Willingness to Pay or the market price. The eco-burden is expressed in eco-costs, a monetised single indicator in LCA (Life Cycle Assessment): an app for IOS and Android, and excel look-up tables at the internet, enable quick assessment of eco-costs. A practical example is given: the design of a new concept of domestic street lighting system for the city of Rotterdam. This new concept results in a considerable reduction of carbon footprint and eco-costs, and shows the benefits for the municipality and for the residents, resulting in a viable business case.

Poincar\xe9 maps for analyzing complex hierarchies in single-cell data

The need to understand cell developmental processes spawned a plethora of computational methods for discovering hierarchies from scRNAseq data. However, existing techniques are based on Euclidean geometry, a suboptimal choice for modeling complex cell trajectories with multiple branches. To overcome this fundamental representation issue we propose Poincaré maps, a method that harness the power of hyperbolic geometry into the realm of single-cell data analysis. Often understood as a continuous extension of trees, hyperbolic geometry enables the embedding of complex hierarchical data in only two dimensions while preserving the pairwise distances between points in the hierarchy. This enables the use of our embeddings in a wide variety of downstream data analysis tasks, such as visualization, clustering, lineage detection and pseudotime inference. When compared to existing methods — unable to address all these important tasks using a single embedding — Poincaré maps produce state-of-the-art two-dimensional representations of cell trajectories on multiple scRNAseq datasets.

Identification of Potential PBT/POP-Like Chemicals by a Deep Learning Approach Based on 2D Structural Features.

Identifying potential persistent organic pollutants (POPs) and persistent, bioaccumulative, and toxic (PBT) substances from industrial chemical inventories are essential for chemical risk assessment, management, and pollution control. Inspired by the connections between chemical structures and their properties, a deep convolutional neural network (DCNN) model was developed to screen potential PBT/POP-like chemicals. For each chemical, a two-dimensional molecular descriptor representation matrix based on 2424 molecular descriptors was used as the model input. The DCNN model was trained via a supervised learning algorithm with 1306 PBT/POP-like chemicals and 9990 chemicals currently known as non-POPs/PBTs. The model can achieve an average prediction accuracy of 95.3 ± 0.6% and an F-measurement of 79.3 ± 2.5% for PBT/POP-like chemicals (positive samples only) on external data sets. The DCNN model was further evaluated with 52 experimentally determined PBT chemicals in the REACH PBT assessment list and correctly recognized 47 chemicals as PBT/non-PBT chemicals. The DCNN model yielded a total of 4011 suspected PBT/POP like chemicals from 58 079 chemicals merged from five published industrial chemical lists. The proportions of PBT/POP-like substances in the chemical inventories were 6.9-7.8%, higher than a previous estimate of 3-5%. Although additional PBT/POP chemicals were identified, no new family of PBT/POP-like chemicals was observed.

Affective Latent Representation of Acoustic and Lexical Features for Emotion Recognition.

In this paper, we propose a novel emotion recognition method based on the underlying emotional characteristics extracted from a conditional adversarial auto-encoder (CAAE), in which both acoustic and lexical features are used as inputs. The acoustic features are generated by calculating statistical functionals of low-level descriptors and by a deep neural network (DNN). These acoustic features are concatenated with three types of lexical features extracted from the text, which are a sparse representation, a distributed representation, and an affective lexicon-based dimensions. Two-dimensional latent representations similar to vectors in the valence-arousal space are obtained by a CAAE, which can be directly mapped into the emotional classes without the need for a sophisticated classifier. In contrast to the previous attempt to a CAAE using only acoustic features, the proposed approach could enhance the performance of the emotion recognition because combined acoustic and lexical features provide enough discriminant power. Experimental results on the Interactive Emotional Dyadic Motion Capture (IEMOCAP) corpus showed that our method outperformed the previously reported best results on the same corpus, achieving 76.72% in the unweighted average recall.

The minimal seesaw model with a modular S4 symmetry

In this paper, we incorporate the modular S4 flavor symmetry into the supersymmetric version of the minimal type-I seesaw model, in which only two right-handed neutrino singlets are introduced to account for tiny Majorana neutrino masses, and explore its implications for the lepton mass spectra, flavor mixing and CP violation. The basic idea is to assign two right-handed neutrino singlets into the unique two-dimensional irreducible representation of the modular S4 symmetry group. Moreover, we show that the matter-antimatter asymmetry in our Universe can be successfully explained via the resonant leptogenesis mechanism working at a relatively-low seesaw scale ΛSS≈ 107 GeV, with which the potential problem of the gravitino overproduction can be avoided. In this connection, we emphasize that the observed matter-antimatter asymmetry may lead to a stringent constraint on the parameter space and testable predictions for low-energy observables.

“The impulse is cartographic”: Counter‐Mapping Indonesia’s Resource Frontiers in the Context of Coloniality

AbstractResource frontiers continue to expand globally across Indigenous lands as states and corporations enact forms of expropriation redolent of the formal colonial era for the sake of extraction. In the face of this expansion, the burden remains largely on frontline communities to defend their ecologies using the tools available to them. Across Indonesia’s resource frontiers, the “cartographic impulse” Edward Said once named to describe anticolonial struggles is apparent in the form of counter‐mapping, which seeks to secure adat (customary) rights and defend Indigenous lands against extractivist expansion. This article revisits this practice and argues that the counter‐map and its goals remain tenuous—mapping in its scalable form risks processing complex and multi‐dimensional ways of relating to land into two‐dimensional representations appropriate for a liberal property regime, while adat itself is a contingent and mutable legal goal. Ultimately, the article echoes emerging calls for the burden to be shifted away from frontline communities through the pursuit of just transitions to post‐extractivism. Pemburuan sumber daya ke wilayah‐wilayah baru yang masih kaya (resource frontiers) di tanah masyarakat adat semakin meluas secara global, seiring praktek‐praktek perampasan berbau kolonial yang dilakukan oleh negara dan korporasi demi tujuan ekstraksi. Berhadapan dengan ekspansi ini, masyarakat di gugus depanlah yang menanggung beban terbesar untuk mempertahankan lingkungan hidup mereka, dengan menggunakan sarana yang tersedia bagi mereka. Di wilayah‐wilayah baru ekstraksi sumber daya alam di Indonesia, apa yang pernah disebut Edward Said sebagai ‘cartographic impulse’ untuk menggambarkan perjuangan anti‐kolonial terlihat jelas dalam bentuk counter‐mapping, yang bertujuan untuk mempertahankan tanah‐tanah adat dari ekspansi ekstravistik. Artikel ini hendak meninjau kembali praktek counter‐mapping ini dan berargumen bahwa counter‐map dan tujuan‐tujuannya masih lemah– mapping dalam bentuk yang dapat diukur mengalami masalah dengan persoalan proses dan persoalan corak multidimensioanal dari hubungan dengan tanah dalam kaitan dengan representasi dua‐dimensi yang hanya cocok dengan rezim kepemilikan liberal; sementara adat sendiri adalah tujuan legal yang sifatnya kontingen dan terbuka terhadap perubahan. Pada akhirnya artikel ini ikut menyuarakan seruan yang banyak muncul, bahwa beban harus dialihkan dari masyarakat frontline melalui usaha untuk transisi‐transisi yang adil menuju post‐extractivism.

Dogs (Canis familiaris) recognise our faces in photographs: implications for existing and future research.

Dogs are an ideal species to investigate phylogenetic and ontogenetic factors contributing to face recognition. Previous research has found that dogs can recognise their owner using visual information about the person's face, presented live. However, a thorough investigation of face processing mechanisms requires the use of graphical representations and it currently remains unclear whether dogs are able to spontaneously recognise human faces in photographs. To test this, pet dogs (N = 60) were briefly separated from their owners and, to achieve reunion, they needed to select the location indicated by a photograph of their owner's face, rather than that of an unfamiliar person concurrently presented. Photographs were taken under optimal and suboptimal (non-frontally oriented and unevenly illuminated faces) conditions. Results revealed that dogs approached their owner significantly above chance level when presented with photos taken under optimal conditions. Further analysis revealed no difference in the probability of choosing the owner between the optimal and suboptimal conditions. Dogs were more likely to choose the owner if they directed a higher percentage of looking time towards the owner's photograph compared to the stranger's one. In addition, the longer the total viewing time of both photos, the higher the probability that dogs chose the stranger. A main effect of dogs' sex was also obtained, with a higher probability of male dogs choosing the owner's photograph. This study provides direct evidence that dogs are able to recognise their owner's face from photographs. The results imply that motion and three-dimensional information is not necessary for recognition. The findings also support the ecological valence of such stimuli and increase the validity of previous investigations into dog cognition that used two-dimensional representations of faces. The effects of attention may reflect differences at the individual level in attraction towards novel faces or in the recruitment of different face processing mechanisms.

Lexis fields

Lexis surfaces are visualizations designed to show how a given value changes over age and time. Vector fields are two-dimensional representations of two variables: usually direction and speed (or force).

Cone beamed computed tomography in pediatric dentistry: Concepts revisited

Oral diagnosis and treatment planning is of utmost importance in pediatric dentistry. Although intraoral and conventional radiographic procedures have been used extensively since decades, there two-dimensional representation has raised many questions. Cone beamed computed tomography with 3-D technology is a replacement for conventional 2-D imaging and has a wide application among child patients in pediatric dentistry. This article therefore deals with current facts and myths about clinical situations where CBCT is indicated in pediatric patients.

The Use of Biometrics and Rubric Analysis to Assess How Students View, Cognitively Process, and Respond to 3D Physical Molecular Models in Undergraduate Biochemistry

Among the STEM disciplines biochemistry curricula present a particular challenge to students with abstract concepts that can lead to rampant and resistant misconceptions impeding learning and therefore impact retention. Moreover, research has shown two‐dimensional representations not only fail to effectively convey biochemical concepts, but also propagate misconceptions. In particular, research has shown students have significant difficulty understanding enzyme‐substrate relationships. One way instructors have suggested to support students while learning this content is through the use of 3D physical models of the enzyme and substrate. However, little research has been conducted on how the models impact the learning process and how to best design these models and corresponding activities. This presentation will describe an iterative design process based on biometric, observational and rubric data analysis. The biometric data is collected using electroencephalographic (EEG) and eye tracking tools and voice recordings from simulated learning environments where biochemistry students completed a serine protease activity with corresponding physical models. These results offer insights into how students view and cognitively process activities using physical models with respect to how the instructor intends students to learn with and use these tools. In addition, observational and rubric analyses of student responses to activities from the simulated learning environment and real classroom sessions provide further data for the iterative activity design process.Support or Funding InformationThis project is supported by the National Science Foundation under award number IUSE 1711402/1711425 to University of Minnesota, Rochester and Kennesaw State University.

The Impact of Parameterized Lateral Mixing on the Antarctic Circumpolar Current in a Coupled Climate Model

AbstractThis study examines the impact of changing the lateral diffusion coefficient ARedi on the transport of the Antarctic Circumpolar Current (ACC). The lateral diffusion coefficient ARedi is poorly constrained, with values ranging across an order of magnitude in climate models. The ACC is difficult to accurately simulate, and there is a large spread in eastward transport in the Southern Ocean (SO) in these models. This paper examines how much of that spread can be attributed to different eddy parameterization coefficients. A coarse-resolution, fully coupled model suite was run with ARedi = 400, 800, 1200, and 2400 m2 s−1. Additionally, two simulations were run with two-dimensional representations of the mixing coefficient based on satellite altimetry. Relative to the 400 m2 s−1 case, the 2400 m2 s−1 case exhibits 1) an 11% decrease in average wind stress from 50° to 65°S, 2) a 20% decrease in zonally averaged eastward transport in the SO, and 3) a 14% weaker transport through the Drake Passage. The decrease in transport is well explained by changes in the thermal current shear, largely due to increases in ocean density occurring on the northern side of the ACC. In intermediate waters these increases are associated with changes in the formation of intermediate waters in the North Pacific. We hypothesize that the deep increases are associated with changes in the wind stress curl allowing Antarctic Bottom Water to escape and flow northward.

Human Action Recognition Based on a Spatio-Temporal Video Autoencoder

Due to rapid advances in the development of surveillance cameras with high sampling rates, low cost, small size and high resolution, video-based action recognition systems have become more commonly used in various computer vision applications. Human operators can be supported with the aid of such systems to detect events of interest in video sequences, improving recognition results and reducing failure cases. In this work, we propose and evaluate a method to learn two-dimensional (2D) representations from video sequences based on an autoencoder framework. Spatial and temporal information is explored through a multi-stream convolutional neural network in the context of human action recognition. Experimental results on the challenging UCF101 and HMDB51 datasets demonstrate that our representation is capable of achieving competitive accuracy rates when compared to other approaches available in the literature.

The Networks Infection Contagious Diseases Positioning System (NICDP-System): The Case of Wuhan-COVID-19

This paper formulates an analytical framework to understand the spatiotemporal patterns of epidemic disease occurrence, its relevance and implications to economic activity, their impact on the vulnerability of different populations, and to develop public health policy decisions on disease prevention issues. The paper suggests a paradigm shift: a new multidimensional geometric approach to capture all symmetrical and asymmetrical strategic graphical movements in real space and time. The Networks Infection Contagious Diseases Positioning System (NICDP-System) extends the traditional two-dimensional graphical representation into a multidimensional graphical framework that synchronizes the optimal virus behavior in real-time. The model investigates the epidemic 2019 novel coronavirus (COVID-19) identified in Wuhan City, Hubei Province, China.

Proximal Over-resection During Femoral Osteochondroplasty Negatively Affects The Distractive Stability Of The Hip Joint

Objectives: Hip microinstability is an increasingly recognized source of hip pain and disability. Femoral osteochondroplasty is usually performed with direct visualization through an arthroscope, assisted with repeated fluoroscopic imaging. However, a two-dimensional representation of a three-dimensional structure is misleading and may compromise the precision of the planned osteochondroplasty. The resection can occasionally extend proximally into the Femoral Head diminishing the articular surface area available for suction seal. The purpose of this study was to determine whether proximal over-resection decreases the rotational and distractive stability of the hip joint. Methods: Six hemi-pelvises were repeatedly tested in the following conditions: (a) intact, (b) T-capsulotomy, (c) Osteochondroplasty to the physeal scar, followed by (d) 5mm and (e) 10mm proximal extension. The pelvis was secured to a metal plate and the femur was potted and attached to a multi-axial hip jig. Specimens were axially distracted with a load from 0-150N followed by 5Nm of internal and external torque at 0o, 15o, 30o, 60o, 90o of flexion. Displacement/rotation was recorded using a 3D motion tracking system. Repeated measures ANOVA was used with significance set at p&lt;0.05. Results: Proximal extension of the resection by 5mm and 10mm increased axial instability at every angle of flexion tested, with the greatest increase observed at higher angles of flexion,p&lt;0.05. T-capsulotomy alone increased both internal and external rotation at all angles of flexion, p&lt;0.05. Subsequent resection and extension of the resection did not significantly increase rotational instability compared to the capsulotomy state. Conclusion: Extending the osteochondroplasty proximally into the femoral head compromises the distractive stability of the hip joint, but doesn’t not effect hip rotational stability. Clinically, this highlights the importance of accuracy when performing femoral osteochondroplasty to minimize proximal extension that may increase iatrogenic instability of the hip joint, leading to reduced post-operative outcomes.

Adaptive structure modeling and prediction for swarm unmanned system

In recent years, swarm unmanned systems (SUSs) have become crucial in the military field, both at home and abroad. This has promoted the evolution of the unmanned combat mode from single-platform remote-control to intelligent-swarm combat. SUSs support the cooperative, autonomous, and flexible characteristics of the combat system under uncertain tasks and environments. The overall swarm performance depends on the system and structure among its members and also dynamically evolves with the time and the environment. Thus, new intelligence emerges from the interaction among systems. Starting from the evolution of the SUS structure, this paper proposes the model of a three-layer structure and a relationship involving the data-link layer, the SUS, and the task requirements. The multidimensional spatial relationship model is transformed into a two-dimensional graphical representation model by using a graph neural network; then, the dependency graph of the relationships of the systems and layers of the SUS is constructed. The integral network is classified according to a task-based standard. The recursive neural network algorithm is derived from the intra- and interlayer relationships. The SUS structure is predicted via some examples of training data sets and the attribute labels of task-based nodes. The impact of the system or data layer damage can be evaluated according to the weight parameters of the structure dependence relationship. Finally, the autonomous decision of the SUS from the task to the swarm structure is realized.

Testing Distinct Three-Dimensional Effects in Laparoscopy: A Prospective Randomized Trial Using the Lübecker Toolbox Curriculum

Background: The use of stereoscopic laparoscopic systems in minimally invasive surgery (MIS) allows a three-dimensional (3D) view of the surgical field, which improves hand-eye coordination. Depending on the stereo base used in the construction of the endoscopes, 3D systems may differ regarding the 3D effect. Our aim was to investigate the influence of different stereo bases on the 3D effect. Methods: This was a prospective randomized study involving 42 MIS-inexperienced study participants. We evaluated two laparoscopic 3D systems with stereo bases of 2.5 mm (system A) and 3.8 mm (system B) for differences in learning MIS skills using the Lübeck Toolbox (LTB) video box trainer. We evaluated participants’ performance regarding the times and repetitions required to reach each exercise’s goal. After completing the final exercise (“suturing”), participants performed the exercise again using a two-dimensional (2D) representation. Additionally, we retrospectively compared our study results with a preliminary study from participants completing the LTB curriculum with a 2D system. Results: The median number of repetitions until reaching the goals for LTB exercises 1, 2, 3, and 6 for system A were: 18 (range 7–53), 24 (range 8–46), 24 (range 13–51), and 21 (range 10–46), respectively, and for system B were: 12 (range 2–30), 16 (range 6–43), 17 (range 4–47), and 15 (range 6–29), respectively (p = not significant). Changing from a 3D to a 2D representation after completing the learning curve led to a longer average time required, from 95.22 to 119.3 s (p < 0.0001), for the last exercise (exercise 6; “suturing”). When comparing the results retrospectively with the learning curves acquired with the 2D system, there was a significant reduction in the number of repetitions required to reach the LTB exercise goals for exercises 1, 3, and 6 using the 3D system. Conclusion: Stereo bases of 2.5 and 3.8 mm provide acceptable bases for designing 3D systems. Additionally, our results indicated that MIS basic skills can be learned quicker using a 3D system versus a 2D system, and that when the 3D effect is eliminated, the corresponding compensatory mechanisms must be relearned.

Constructions of vector-valued modular forms of rank four and level one

This paper studies modular forms of rank four and level one. There are two possibilities for the isomorphism type of the space of modular forms that can arise from an irreducible representation of the modular group of rank four, and we describe when each case occurs for general choices of exponents for the [Formula: see text]-matrix. In the remaining sections we describe how to write down the corresponding differential equations satisfied by minimal weight forms, and how to use these minimal weight forms to describe the entire graded module of holomorphic modular forms. Unfortunately, the differential equations that arise can only be solved recursively in general. We conclude the paper by studying the cases of tensor products of two-dimensional representations, symmetric cubes of two-dimensional representations, and inductions of two-dimensional representations of the subgroup of the modular group of index two. In these cases, the differential equations satisfied by minimal weight forms can be solved exactly.

Predicting drug–protein interaction using quasi-visual question answering system

Identifying novel drug–protein interactions is crucial for drug discovery. For this purpose, many machine learning-based methods have been developed based on drug descriptors and one-dimensional protein sequences. However, protein sequences cannot accurately reflect the interactions in three-dimensional space. However, direct input of three-dimensional structure is of low efficiency due to the sparse three-dimensional matrix, and is also prevented by the limited number of co-crystal structures available for training. Here we propose an end-to-end deep learning framework to predict the interactions by representing proteins with a two-dimensional distance map from monomer structures (Image) and drugs with molecular linear notation (String), following the visual question answering mode. For efficient training of the system, we introduce a dynamic attentive convolutional neural network to learn fixed-size representations from the variable-length distance maps and a self-attentional sequential model to automatically extract semantic features from the linear notations. Extensive experiments demonstrate that our model obtains competitive performance against state-of-the-art baselines on the directory of useful decoys, enhanced (DUD-E), human and BindingDB benchmark datasets. Further attention visualization provides biological interpretation to depict highlighted regions of both protein and drug molecules. When predicting the interaction of proteins with potential drugs, the protein can be encoded as its one-dimensional sequence or a three-dimensional structure, which can capture more relevant features of the protein, but also makes the task to predict the interactions harder. A new method predicts these interactions using a two-dimensional distance matrix representation of a protein, which can be processed like a two-dimensional image, striking a balance between the data being simple to process and rich in relevant structures.