Symmetric Alignment Research Articles

Multi-level Symmetric Semantic Alignment Network for image–text matching

Image–text matching has attracted much attention as one of the visual-linguistic tasks. Most of the existing methods tend to concentrate on single-level semantic similarity by global embeddings or local alignment, which fail to distinguish the ambiguous image–text instances with highly similar contexts due to insufficient interactions of cross-modal features. To mitigate this problem, in this paper, we propose a novel Multi-level Symmetric Semantic Alignment Network (MSSAN), which fully exploits the multi-level semantic features to capture the complicated correlations between images and text. Specifically, global–global-level alignment is first performed based on visual–textual global representations. Then, considering the local semantic consistency, local–local-level alignment is carried out through inter-modal bidirectional cross-attention module to model fine-grained region–word relations. Moreover, a Multi-granularity Feature Fusion Module (MFFM) is constructed to learn intra-modal distribution of significance and incorporate global concepts into local features to obtain more comprehensive semantic representations, thus achieving global–local-level alignment. Finally, in order to achieve a strong separation of semantically similar samples, we develop a novel generic triplet ranking loss with adaptively updatable margins to train our model. Extensive experimental results on Flickr8K, MSCOCO and Flickr30K datasets demonstrate that our proposed MSSAN is superior to other state-of-the-art methods by a considerable margin.

Tablet-based Augmented reality and 3D printed templates in fully guided Microtia Reconstruction: a clinical workflow

BackgroundMicrotia is a congenital malformation of the auricle that affects approximately 4 of every 10,000 live newborns. Radiographic film paper is traditionally employed to bidimensionally trace the structures of the contralateral healthy ear in a quasi-artistic manner. Anatomical points provide linear and angular measurements. However, this technique proves time-consuming, subjectivity-rich, and greatly dependent on surgeon expertise. Hence, it’s susceptible to shape errors and misplacement.MethodsWe present an innovative clinical workflow that combines 3D printing and augmented reality (AR) to increase objectivity and reproducibility of these procedures. Specifically, we introduce patient-specific 3D cutting templates and remodeling molds to carve and construct the cartilaginous framework that will conform the new ear. Moreover, we developed an in-house AR application compatible with any commercial Android tablet. It precisely guides the positioning of the new ear during surgery, ensuring symmetrical alignment with the healthy one and avoiding time-consuming intraoperative linear or angular measurements. Our solution was evaluated in one case, first with controlled experiments in a simulation scenario and finally during surgery.ResultsOverall, the ears placed in the simulation scenario had a mean absolute deviation of 2.2 ± 1.7 mm with respect to the reference plan. During the surgical intervention, the reconstructed ear was 3.1 mm longer and 1.3 mm wider with respect to the ideal plan and had a positioning error of 2.7 ± 2.4 mm relative to the contralateral side. Note that in this case, additional morphometric variations were induced from inflammation and other issues intended to be addressed in a subsequent stage of surgery, which are independent of our proposed solution.ConclusionsIn this work we propose an innovative workflow that combines 3D printing and AR to improve ear reconstruction and positioning in microtia correction procedures. Our implementation in the surgical workflow showed good accuracy, empowering surgeons to attain consistent and objective outcomes.

Automatic Texture Alignment by Optimization Method.

Microstructure analysis via electron backscatter diffraction has become an indispensable tool in materials science and engineering. In order to interpret or predict the anisotropy in crystalline materials, the texture is assessed, e.g. via pole figure diagrams. To ensure a correct characterization, it is crucial to align the measured sample axes as closely as possible with the manufacturing process directions. However, deviations are inevitable due to sample preparation and manual measurement setup. Postprocessing is mostly done manually, which is tedious and operator-dependent. In this work, it is shown that the deviation can be calculated using the contour of the crystal orientations. This can also be utilized to define the axis symmetry of pole figure diagrams through an objective function, allowing for symmetric alignment by minimization. Experimental textures of extruded profiles and synthetically generated textures were used to demonstrate the general applicability of the method. It has proven to work excellently for deviations of up to 5∘, which are typical for careful manual sample preparation and mounting. While the performance of the algorithm is reduced with increasing misalignment, good results have also been obtained for deviations up to 15∘.

The basis invariant flavor puzzle

The flavor puzzle of the Standard Model quark sector is formulated in a non-perturbative way, using basis invariants that are independent of the choice of quark field basis. To achieve this, we first derive the algebraic ring of 10 CP even (primary) and 1 CP odd (secondary) basis invariants, using the Hilbert series and plethystic logarithm. An orthogonal basis in the ring of basis invariants is explicitly constructed, using hermitian projection operators derived via birdtrack diagrams. The thereby constructed invariants have well defined CP transformation behavior and give the most direct access to the flavor symmetric alignments of basis covariants. We firstly “measure” the orthogonal basis invariants from experimental data and characterize their location in the available parameter space. The experimentally observed orthogonal basis invariants take very close to maximal values and are highly correlated. Explaining the location of the invariants at close to maximal points, including the associated miniscule and highly correlated deviations, corresponds to solving the flavor puzzle in the invariant language. Once properly normalized, the orthogonal basis invariants are close to scale (RGE) invariant, hence, provide exquisite targets for fits of both, low- and high-scale (bottom-up and top-down) flavor models. Our result provides an entirely new angle on the flavor puzzle, and opens up ample opportunities for its ultimate exploration.

Benefits of negative pressure wound therapy with fat migration during revisional total hip arthroplasty in an obese patient: A case report

Introduction: A 38-year-old African American woman presented with right hip pain and movement restriction. Her medical history included a right hip and knee arthroplasty 10 years prior, history of Slipped Capital Femoral Epiphysis and osteonecrosis of the hip. Preoperative assessment was significant for multiple comorbidities such as obesity (BMI > 38), hypertension, asthma. Patient concerns: The patient presented with right hip pain, rated 7/10, and restricted hip flexion, adduction, and abduction. Diagnosis: Recent imaging showed eccentric deterioration of the polyethylene lining of her prosthesis, acetabular hypertrophy on her right hip prosthesis, and chronic deformity of the pubic bone. Interventions: Based on these findings, a revisional total hip arthroplasty was performed. After the surgical procedure, the WoundVAC and the percutaneous drain were applied outside the tensor fascia lata to reduce seroma and hematoma formation. Postoperative pain control, antibiotics and DVT prophylaxis were given. On post-op day 3, an irrigation and debridement with delayed primary wound closure was performed under sterile conditions. Outcomes: On postoperative assessment, the wound demonstrated adequate healing without any signs of infection. Sutures and staples were removed 4 weeks post-op. Upon palpation there was no edema, effusions, temperature changes, tenderness. Clinical inspection revealed symmetrical alignment of the pelvis and hips. Range of motion testing revealed restriction beyond 80 degrees upon flexion and beyond 5 degrees of adduction and 10 degrees of abduction. The surgical site was noted to be healed at 6 weeks post-op. The patient continued to do well to date, without exacerbations. Conclusion: Obesity increases the risk of post-operative complications and wound healing failure. Therefore, Vacuum-assisted wound closure (WoundVAC), a type of negative pressure wound therapy, was applied outside the tensor fascia lata post-operatively, where the surgical incision was made. Negative pressure wound therapy facilitates wound healing by stimulating angiogenesis and promoting granulation tissue formation, which in turn can reduce the risk of surgical site infection in obese patients undergoing total hip arthroplasty. Highlighted is the mechanism of fat migration in the promotion of wound healing after preoperative weight loss and exercise.

吉林省 和龍市 渤海 北大古墳群 考察

The Beida Ancient Tombs, situated in Baijazi, Helong, Jilin Province, is on the northern bank of the Hailan River, approximately 5 km southwest of Xigucheng. The tombs primarily exhibit a lateral crypt (vaulted) burial style with stone chambers and enclosures. However, some instances of stone-walled earthen mounds have been identified.
 The interred individuals were either buried in wooden coffins or directly into a platform. While there is evidence of structures for placing a coffin, commonly, the natural ground surface was utilized without alteration. Meanwhile, within individual tombs, additional burials took place over a long period. It is noteworthy that individuals were laid to rest in different ways. While artifact excavation was not abundant due to simple funerals, the central chronological period has been ascertained to span from the pre-8th century to the mid of that century through cross-dating methods involving Chinese ceramics, bronze mirrors, and Tang-style metal belts.
 The Beida Ancient Tombs are predominantly stone-built and the prevalence of excavated foreign artifacts is the highest among the Balhae tombs. In particular, given its symmetrical alignment with the Longtoushan graveyard, particularly centered around Xigucheng, the construction of the Beida Ancient Tombs was influenced by the funerary grounds of officials closely connected to the central capital of Balhae Jungkyung hyundeokbu.

Unsupervised deep consistency learning adaptation network for cardiac cross-modality structural segmentation.

Deep neural networks have recently been succeessful in the field of medical image segmentation; however, they are typically subject to performance degradation problems when well-trained models are tested in another new domain with different data distributions. Given that annotated cross-domain images may inaccessible, unsupervised domain adaptation methods that transfer learnable information from annotated source domains to unannotated target domains with different distributions have attracted substantial attention. Many methods leverage image-level or pixel-level translation networks to align domain-invariant information and mitigate domain shift issues. However, These methods rarely perform well when there is a large domain gap. A new unsupervised deep consistency learning adaptation network, which adopts input space consistency learning and output space consistency learning to realize unsupervised domain adaptation and cardiac structural segmentation, is introduced in this paper The framework mainly includes a domain translation path and a cross-modality segmentation path. In domain translation path, a symmetric alignment generator network with attention to cross-modality features and anatomy is introduced to align bidirectional domain features. In the segmentation path, entropy map minimization, output probability map minimization and segmentation prediction minimization are leveraged to align the output space features. The model conducts supervised learning to extract source domain features and conducts unsupervised deep consistency learning to extract target domain features. Through experimental testing on two challenging cross-modality segmentation tasks, our method has robust performance compared to that of previous methods. Furthermore, ablation experiments are conducted to confirm the effectiveness of our framework.

General and Reliable Azimuthal Alignment Algorithm for Low Loss Multicore Fiber Fusion Splicing

Multicore fiber is a promising spatial division multiplexing technology to increase the transmission bandwidth of optical fiber links in any given available spatial cross section. To connect multicore fibers, similar to single-core fibers, fusion splicing is an indispensable tool. Since multicore fibers cannot be circularly symmetric, precise transverse and azimuthal alignment is crucial to minimize the insertion loss penalty that would originate from even small offsets in the core locations of the two fibers to be spliced together. In this paper, a novel azimuthal alignment algorithm for multicore fiber splicing is presented. Since it is based on side-view images, it can also determine the twist rates of both fibers to be spliced and take these twist rates into account when calculating the optimum azimuthal alignment angle at the very ends of the two fibers. In the case of fibers with markers, i.e., fibers that are not rotationally symmetric, the relative and absolute polarities of the two fiber ends are computed as well. We present the formulation of the azimuthal alignment algorithm and demonstrate its stability and versatility for a wide range of fibers with vastly different geometries, core numbers and twist rates, including an offset single-core fiber. In the case of 4-core fibers with markers, we achieve average splice losses of less than 0.03 dB regardless of the polarity, which is a substantial advantage in terms of real-world applicability in the field. For a twisted 7-core fiber, we demonstrate an accurate twist rate detection with standard deviation < 1.4/m (turns per meter).

Multi‐agent Path Planning with Heterogenous Interactions in Tight Spaces

AbstractBy starting with the assumption that motion is fundamentally a decision making problem, we use the world‐line concept from Special Relativity as the inspiration for a novel multi‐agent path planning method. We have identified a particular set of problems that have so far been overlooked by previous works. We present our solution for the global path planning problem for each agent and ensure smooth local collision avoidance for each pair of agents in the scene. We accomplish this by modelling the collision‐free trajectories of the agents through 2D space and time as rods in 3D. We obtain smooth trajectories by solving a non‐linear optimization problem with a quasi‐Newton interior point solver, initializing the solver with a non‐intersecting configuration from a modified Dijkstra's algorithm. This space–time formulation allows us to simulate previously ignored phenomena such as highly heterogeneous interactions in very constrained environments. It also provides a solution for scenes with unnaturally symmetric agent alignments without the need for jittering agent positions or velocities.

Patterned proliferation orients tissue-wide stress to control root vascular symmetry in Arabidopsis

Symmetric tissue alignment is pivotal to the functions of plant vascular tissue, such as long-distance molecular transport and lateral organ formation. During the vascular development of the Arabidopsis roots, cytokinins initially determine cell-type boundaries among vascular stem cells and subsequently promote cell proliferation to establish vascular tissue symmetry. Although it is unknown whether and how the symmetry of initially defined boundaries is progressively refined under tissue growth in plants, such boundary shapes in animal tissues are regulated by cell fluidity, e.g., cell migration and intercalation, lacking in plant tissues. Here, we uncover that cell proliferation during vascular development produces anisotropic compressive stress, smoothing, and symmetrizing cell arrangement of the vascular-cell-type boundary. Mechanistically, the GATA transcription factor HANABA-TARANU cooperates with the type-B Arabidopsis response regulators to form an incoherent feedforward loop in cytokinin signaling. The incoherent feedforward loop fine-tunes the position and frequency of vascular cell proliferation, which in turn restricts the source of mechanical stress to the position distal and symmetric to the boundary. By combinatorial analyses of mechanical simulations and laser cell ablation, we show that the spatially constrained environment of vascular tissue efficiently entrains the stress orientation among the cells to produce a tissue-wide stress field. Together, our data indicate that the localized proliferation regulated by the cytokinin signaling circuit is decoded into a globally oriented mechanical stress to shape the vascular tissue symmetry, representing a reasonable mechanism controlling the boundary alignment and symmetry in tissue lacking cell fluidity.

Discrete Orientations of Interfacial Waters Direct Crystallization of Mica-Binding Proteins.

Understanding the basis of templated molecular assembly on a solid surface requires a fundamental comprehension of both short- and long-range aqueous response to the surface under a variety of solution conditions. Herein we provide a detailed picture of how the molecular-scale response to different mica surfaces yields distinct solvent orientations that produce quasi-static directional potentials onto which macromolecules can adsorb. We connect this directionality to observed (a)symmetric epitaxial alignment of designed proteins onto these surfaces, corroborate our findings with 3D atomic force microscopy experiments, and identify slight differences in surface structure as the origin of this effect. Our work provides a detailed picture of the intrinsic electrolyte response in the vicinity of mineral interfaces, with clear predictions for experiment, and highlights the role of solvent on the predictive assembly of hierarchical materials on mineral surfaces.

Identify Bitter Peptides by Using Deep Representation Learning Features.

A bitter taste often identifies hazardous compounds and it is generally avoided by most animals and humans. Bitterness of hydrolyzed proteins is caused by the presence of bitter peptides. To improve palatability, bitter peptides need to be identified experimentally in a time-consuming and expensive process, before they can be removed or degraded. Here, we report the development of a machine learning prediction method, iBitter-DRLF, which is based on a deep learning pre-trained neural network feature extraction method. It uses three sequence embedding techniques, soft symmetric alignment (SSA), unified representation (UniRep), and bidirectional long short-term memory (BiLSTM). These were initially combined into various machine learning algorithms to build several models. After optimization, the combined features of UniRep and BiLSTM were finally selected, and the model was built in combination with a light gradient boosting machine (LGBM). The results showed that the use of deep representation learning greatly improves the ability of the model to identify bitter peptides, achieving accurate prediction based on peptide sequence data alone. By helping to identify bitter peptides, iBitter-DRLF can help research into improving the palatability of peptide therapeutics and dietary supplements in the future. A webserver is available, too.

Room temperature ferromagnetism and transport properties in InN/VTe2 van der Waals heterostructures

The room temperature ferromagnetism can be induced for nonmagnetic pristine graphene-like InN monolayer through establishing van der Waals heterostructure with ferromagnetic semiconductor, which could be applied for realizing novel low-dimensional spintronic devices. Here, we investigate the electronic structure and magnetic properties of InN/VTe2 van der Waals heterostructure under strain and electrostatic doping by first-principles calculations. The InN/VTe2 van der Waals heterostructure possesses desired physical properties such as a symmetric band alignment, ferromagnetic ground state, large in-plane magnetic anisotropy energy of 1.54 meV, and high Curie temperature of 354 K. Meanwhile, the band alignment and magnetic anisotropy energy for InN/VTe2 van der Waals heterostructure can be also tuned effectively by strain. Significant enhancement of conductivity and robust above room temperature ferromagnetism exhibit in InN/VTe2 van der Waals heterostructure under electrostatic doping and paves a new route toward a low-dimensional spintronic field-effect transistor.

Bilaterally Asymmetric Helical Myofibrils in Ascidian Tadpole Larvae.

The locomotor system is highly bilateral at the macroscopic level. Homochirality of biological molecules is fully compatible with the bilateral body. However, whether and how single-handed cells contribute to the bilateral locomotor system is obscure. Here, exploiting the small number of cells in the swimming tadpole larva of the ascidian Ciona, we analyzed morphology of the tail at cellular and subcellular scales. Quantitative phase-contrast X-ray tomographic microscopy revealed a high-density midline structure ventral to the notochord in the tail. Muscle cell nuclei on each side of the notochord were roughly bilaterally aligned. However, fluorescence microscopy detected left-right asymmetry of myofibril inclination relative to the longitudinal axis of the tail. Zernike phase-contrast X-ray tomographic microscopy revealed the presence of left-handed helices of myofibrils in muscle cells on both sides. Therefore, the locomotor system of ascidian larvae harbors symmetry-breaking left-handed helical cells, while maintaining bilaterally symmetrical cell alignment. These results suggest that bilateral animals can override cellular homochirality to generate the bilateral locomotor systems at the supracellular scale.

A Novel (2, 3)-Threshold Reversible Secret Image Sharing Scheme Based on Optimized Crystal-Lattice Matrix

The (k, n)-threshold reversible secret image sharing (RSIS) is technology that conceals the secret data in a cover image and produces n shadow versions. While k (k ≤ n) or more shadows are gathered, the embedded secret data and the cover image can be retrieved without any error. This article proposes an optimal (2, 3) RSIS algorithm based on a crystal-lattice matrix. Sized by the assigned embedding capacity, a crystal-lattice model is first generated by simulating the crystal growth phenomenon with a greedy algorithm. A three-dimensional (3D) reference matrix based on translationally symmetric alignment of crystal-lattice models is constructed to guide production of the three secret image shadows. Any two of the three different shares can cooperate to restore the secret data and the cover image. When all three image shares are available, the third share can be applied to authenticate the obtained image shares. Experimental results prove that the proposed scheme can produce secret image shares with a better visual quality than other related works.

Anticancer peptides prediction with deep representation learning features.

Anticancer peptides constitute one of the most promising therapeutic agents for combating common human cancers. Using wet experiments to verify whether a peptide displays anticancer characteristics is time-consuming and costly. Hence, in this study, we proposed a computational method named identify anticancer peptides via deep representation learning features (iACP-DRLF) using light gradient boosting machine algorithm and deep representation learning features. Two kinds of sequence embedding technologies were used, namely soft symmetric alignment embedding and unified representation (UniRep) embedding, both of which involved deep neural network models based on long short-term memory networks and their derived networks. The results showed that the use of deep representation learning features greatly improved the capability of the models to discriminate anticancer peptides from other peptides. Also, UMAP (uniform manifold approximation and projection for dimension reduction) and SHAP (shapley additive explanations) analysis proved that UniRep have an advantage over other features for anticancer peptide identification. The python script and pretrained models could be downloaded from https://github.com/zhibinlv/iACP-DRLF or from http://public.aibiochem.net/iACP-DRLF/.

Standard reference values of the upper body posture in healthy middle-aged female adults in Germany

In order to classify and analyze the parameters of upper body posture, a baseline in form of standard values is demanded. To this date, standard values have only been published for healthy young women. Data for female adults between 51 and 60 years are lacking. 101 symptom-free female volunteers aged 51–60 (55.16 ± 2.89) years. The mean height of the volunteers was 1.66 ± 0.62 m, with a mean body weight of 69.3 ± 11.88 kg and an average BMI of 25.02 ± 4.55 kg/m2. By means of video raster stereography, a 3D-scan of the upper back surface was measured in a habitual standing position. The confidence interval, tolerance range and ICCs were calculated for all parameters. The habitual standing position is almost symmetrical in the frontal plane the most prominent deviation being a slightly more ventral position of the left shoulder blade in comparison to the right. The upper body (spine position) is inclined ventrally with a minor tilt to the left. In the sagittal plane, the kyphosis angle of the thoracic spine is greater than the lordosis angle of the lumbar spine. The pelvis is virtually evenly balanced with deviations from an ideal position falling under the measurement error margin of 1 mm/1°. There were also BMI influenced postural variations in the sagittal plane and shoulder distance. The ICCs are calculated from three repeated measurements and all parameters can be classified as "almost perfect". Deflections from an ideally symmetric spinal alignment in women aged 51–60 years are small-scaled, with a minimal frontal-left inclination and accentuated sigmoidal shape of the spine. Postural parameters presented in this survey allow for comparisons with other studies as well as the evaluation of clinical diagnostics and applications.

Topological defects arrays and control of electro-convections in periodically photo-aligned bent-core nematics

The applicability of photo-alignment technique to bent-core nematics (BCN) is demonstrated in this paper. In order to cope with the inherent problems of BCN such as high working temperature and large working current, a photo-alignment technique, based on two superimposed photosensitive layers of azo and reactive mesogen, is used to produce symmetric and asymmetric planar alignments. With various surface conditions, defect lines arrays are self-assembled and ac-driven electro-convection (EC) rolls are modulated. Further inspections indicate that the defect lines arrays and EC rolls are in a competitive state, which reveal that the EC rolls in BCN belong to nonstandard EC scenarios. By producing defect lines and modulating EC rolls via photo-alignment technique, we could customize patterns in liquid crystals towards new photonic devices.

Impact of Symmetric Vertical Sinusoid Alignments on Infrastructure Construction Costs: Optimizing Energy Consumption in Metropolitan Railway Lines Using Artificial Neural Networks

Minimizing energy consumption is a key issue from both an environmental and economic perspectives for railways systems; however, it is also important to reduce infrastructure construction costs. In the present work, an artificial neural network (ANN) was trained to estimate the energy consumption of a metropolitan railway line. This ANN was used to test hypothetical vertical alignments scenarios, proving that symmetric vertical sinusoid alignments (SVSA) can reduce energy consumption by up to 18.4% compared with a flat alignment. Finally, we analyzed the impact of SVSA application on infrastructure construction costs, considering different scenarios based on top–down excavation methods. When balancing reduction in energy consumption against infrastructure construction costs between SVSA and flat alignment, the extra construction costs due to SVSA have a return period of 25–300 years compared with a flat alignment, depending on the soil type and construction method used. Symmetric vertical sinusoid alignment layouts are thus suitable for scattered or soft soils, up to compacted intermediate geomaterials.

Word.alignment: an R package for computing statistical word alignment and its evaluation

Word alignment has lots of applications in various natural language processing (NLP) tasks. As far as we are aware, there is no word alignment package in the R environment. In this paper, word.alignment, a new R software package is introduced which implements a statistical word alignment model as an unsupervised learning. It uses IBM Model 1 as a machine translation model based on the use of the EM algorithm and the Viterbi search in order to find the best alignment. It also provides the symmetric alignment using three heuristic methods such as union, intersection, and grow-diag. It has also the ability to build an automatic bilingual dictionary applying an innovative rule. The generated dictionary is suitable for a number of NLP tasks. This package provides functions for measuring the quality of the word alignment via comparing the alignment with a gold standard alignment based on five metrics as well. It is easily installed and executable on the mostly widely used platforms. Note that it is easily usable and we show that its results are almost everywhere better than some other word alignment tools. Finally, some examples illustrating the use of word.alignment is provided.