Search

Traditional machining is transforming to digital and intelligent machining, in which adaptive machining cyber-physical system (CPS) provides a useful approach to control the machining quality of large thin-walled parts. And the running of adaptive machining CPS is a complex multi-processes execution flow, which can be regarded as a continuous–discrete hybrid system. To realise adaptive controlling of machining quality and adaptive managing of process flow, a unified model for continuous–discrete hybrid adaptive machining CPS is constructed. Firstly, an architecture of adaptive machining CPS is proposed. Next, the cutting process in adaptive machining CPS is modelled as a continuous-variable system (CVS), while the process flow in adaptive machining CPS is modelled as a discrete-events system (DES). Then, the finite state machine is adopted to integrate the CVS and DES to form the unified model of adaptive machining CPS. Finally, an adaptive machining CPS is developed based on the unified model, and the machining results demonstrate that machining quality is efficiently controlled, as well as the process flow is orderly managed. The built unified model has four features, respectively universality, integrability, scalability, and reconfigurability, which can be reconstructed to form a new instancing model according to the different machining requirements. [ABSTRACT FROM AUTHOR]

Dynamic analysis of high-speed train–track–bridge (TTB) system presents engineering challenges, particularly under complex conditions like track irregularity and crosswinds. Achieving a balance between computational efficiency and accuracy using the traditional substructure method, i.e. modeling the train and track–bridge separately and possibly performing iterations between them, can be difficult. In this paper, a practical unified three-dimensional (3D) TTB model is presented by using two coordinate systems: a stationary system for the track–bridge, and a moving system for the high-speed train. A wheel–rail interaction (WRI) element model is further extended such that the wheel and rail nodes in WRI element belong to moving and stationary systems, respectively, so as to make the train and track–bridge models consistently combined into one unified model. The unified modeling method simplifies computations and improves efficiency and stability. The modeling method has been implemented into an open source finite element platform, i.e. OpenSees, and applied for dynamic response analysis of a train traveling at 50m/s over a seven-span simply supported bridge. Running safety assessments are performed using three criteria, i.e. the Nadal criterion, Prud’homme criterion and unloading criterion, considering track irregularities and crosswind loads. The analysis results provide practically useful references for engineering. [ABSTRACT FROM AUTHOR]

In this paper, we endeavor to localize all potential objects in an image and infer their visual categories, attributes, and shapes, even in instances where certain objects have not been encompassed in the model's supervised training. This is similar to the challenge posed by open-vocabulary object detection and recognition. The proposed OV-DAR framework, in contrast to previous object detection and recognition frameworks, offers superior advantages and performance in terms of generalization, universality, and granularity expression. Specifically, OV-DAR disentangles the open-vocabulary object detection and recognition problem into two components: class-agnostic object proposal and open-vocabulary classification. It employs co-training to maintain a balance between the performance of these two components. For the former, we construct class-agnostic object proposal networks based on the anchor/query with the SAM foundation model, which demonstrates robust generalization in object proposing and masking. For the latter, we merge available object-centered category classification and attribute prediction data, take co-learning for efficient fine-tuning of CLIP, and subsequently augment the open-vocabulary capability on object-centered category/attribute prediction tasks using freely accessible online image–text pairs. To ensure the efficiency and accuracy of open-vocabulary classification, we devise a structure akin to Faster R-CNN and fully exploit the knowledge of object-centered CLIP for end-to-end multi-object open-vocabulary category and attribute prediction by knowledge distillation. We conduct comprehensive experiments on VAW, MS-COCO, LSA, and OVAD datasets. The results not only illustrate the complementarity of semantic category and attribute recognition for visual scene understanding but also underscore the generalization capability of OV-DAR in localizing, categorizing, attributing, and masking tasks and open-world scene perception. [ABSTRACT FROM AUTHOR]

A virtual synchronous generator (VSG) is prone to transient instability under a grid fault, which leads to the loss of synchronization between the new energy converter and grid, and threatens the operation safety of high-proportion new energy grids. There are a variety of control models in the existing VSG control, including active and reactive power models, which lead to their different transient stabilities. However, the evolution characteristics, correlation between different models of VSG, and the internal mechanism affecting transient stability have not been fully studied. To this effect, this paper analyzes their evolution characteristics based on the existing mainstream VSG control models and establishes a unified VSG model and its equivalent correspondence with other models. Then, the phase plane method is used to comprehensively analyze and compare the transient stability of the VSG unified model with other models. It is pointed out that the key factors affecting the transient stability of different models are three links of primary frequency regulation, reactive power regulation and reactive power tracking. Finally, the correctness of the established VSG unified model and the conclusion of transient stability analysis is verified by experiments. [ABSTRACT FROM AUTHOR]

When the near-field and far-field information of a target is uncertain, it is necessary to choose a suitable localization method. The modified polar representation (MPR) method integrates the two scenarios and achieves a unified localization with direction of arrival (DOA) estimation in the far field and position estimation in the near field. Previous studies have only proposed solutions for stationary environments and have not considered the motion factor. Therefore, this paper proposes a new unified positioning algorithm using multi-sensor time difference of arrival (TDOA) and frequency difference of arrival (FDOA) measurements without prior target source information. The method represents the position of the target source using MPR and describes the localization problem as a weighted least squares (WLS) problem with two constraints. We first obtain the initial estimates by WLS without considering the constraints and then investigate a two-step error correction method based on the constraints. The first step corrects the initial estimate using the Taylor series expansion technique, and the second step corrects the DOA estimate in the previous step using the direct error compensation technique based on the properties of the second constraint. Simulation experiments show that the method is effective for the unified positioning of moving targets and can achieve the Cramer–Rao lower bound (CRLB). [ABSTRACT FROM AUTHOR]

The impact on global simulations of a new package of physical parametrizations in the Met Office Unified Model is documented. The main component of the package is an entirely new convection scheme, CoMorph. This has a mass‐flux structure that allows initiation of buoyant ascent from any level and the ability for plumes of differing originating levels to coexist in a grid box. It has a different form of closure, where the mass flux of initiation is dependent on local instability, and an implicit numerical solution for detrainment that yields smooth timestep behaviour. The scheme is coupled more consistently to the cloud, microphysics, and boundary‐layer parametrizations and, as a result, significant changes to these have also been made. The package, called CoMorph‐A, has been tested in a variety of single‐column and idealized regimes. Here we test it in global configurations and evaluate it against observations using a range of standard metrics. Overall it is found to perform well against the control. Biases in the climatologies of the radiative fluxes are significantly reduced across the Tropics and subtropics, tropical and extratropical cyclone statistics are improved, and the Madden–Julian oscillation and other propagating tropical waves are strengthened. It also improves overall scores in numerical weather prediction trials, without revisions to the data assimilation. There is still work to do to improve the diurnal cycle of precipitation over land, where the peak remains too close to the middle of the day. [ABSTRACT FROM AUTHOR]

In this paper, we present a unified model for gallium arsenide (GaAs) based on ab initio simulations which characterizes its terahertz (THz) properties when excited by optical pump. We use density functional perturbation theory to calculate the dielectric properties of GaAs, and investigate the relaxation time of photo-excited GaAs through electron–phonon interactions. In light of the complexities arising from the mixed absorption mechanisms and the sensitivity of GaAs to laser parameters, we have developed a method that leverages time-dependent density functional theory and Boltzmann transport theory. This approach enables us to establish an accurate relationship between the pump laser intensity and the carrier concentration by introducing the percentage of excited electrons, facilitating the quantitative characterization of GaAs's response under different optical pump intensities. Using the microscopic material parameters solved by first principles, we develop a unified Drude model to describe the macroscopic electromagnetic responses of photo-excited GaAs. We simulate several reported numerical examples of photo-excited GaAs, including a GaAs wafer and GaAs-based THz metamaterial modulators, to validate the proposed unified model as a reliable approach for predicting the THz properties of GaAs. The good agreement between the simulation and measurement results demonstrates that our model successfully captures the dynamic responses of photo-generated carriers and provides guidance for the design of optoelectronic devices based on GaAs. Furthermore, our modeling approach based on ab initio simulations is free from empirical parameters, providing a solid THz modeling method for other photo-excited semiconductor materials. [ABSTRACT FROM AUTHOR]

We investigated positive temperature (warm) and negative precipitation (dry) biases in convection-permitting model (CPM) simulations for Europe (2.2 km grid spacing) that were considerably larger than in equivalent regional climate model (RCM) simulations (12 km grid spacing). We found that improvements in dry biases could be made by (1) using a more complex runoff scheme which takes into account topography and groundwater, (2) delaying the onset of water stress in vegetation to enhance transpiration, (3) changing the microphysics scheme to CASIM (Cloud AeroSol Interacting Microphysics) which also decreases heavy rainfall and increases light rainfall. Increasing soil moisture to the critical point can remove dry precipitation biases in southern Europe but not in northern areas, indicating that soil moisture limitation is a key contributor to precipitation biases in the south only. Instead, in the north, changing the cloud scheme of the model has more impact on precipitation biases. We found that the more intense and intermittent nature of rainfall in the CPM, which is more realistic leads to different canopy interception compared to the RCM. This can impact canopy evaporation, evapotranspiration and feed back on precipitation. Increasing rainfall storage in the canopy only leads to small improvements in warm biases, since it still fills rapidly with intense CPM rainfall, suggesting the need for an additional moisture store via improved groundwater modelling or surface pooling. Overall, this work highlights the challenge of correctly capturing land surface feedbacks in CPMs, which play an important role in future climate projections in some regions. [ABSTRACT FROM AUTHOR]

In this paper, we take the advantage of previous pre-trained models (PTMs) and propose a novel Chinese pre-trained unbalanced transformer (CPT). Different from previous Chinese PTMs, CPT is designed to utilize the shared knowledge between natural language understanding (NLU) and natural language generation (NLG) to boost the performance. CPT consists of three parts: a shared encoder, an understanding decoder, and a generation decoder. Two specific decoders with a shared encoder are pre-trained with masked language modeling (MLM) and denoising auto-encoding (DAE) tasks, respectively. With the partially shared architecture and multi-task pre-training, CPT can (1) learn specific knowledge of both NLU or NLG tasks with two decoders and (2) be fine-tuned flexibly that fully exploits the potential of the model. Moreover, the unbalanced transformer saves the computational and storage cost, which makes CPT competitive and greatly accelerates the inference of text generation. Experimental results on a wide range of Chinese NLU and NLG tasks show the effectiveness of CPT. [ABSTRACT FROM AUTHOR]

Correlating damage outcomes to a retinal laser exposure is critical for diagnosis and choosing appropriate treatment modalities. Therefore, it is important to understand the causal relationships between laser parameters, such as wavelength, power density, and length of exposure, and any resulting injury. Differentiating photothermal from photochemical processes in an in vitro retinal model using cultured retinal pigment epithelial cells would be a first step in achieving this goal. The first-order rate constant of Arrhenius has been used for decades to approximate cellular thermal damage. A modification of this equation, called the damage integral (Ω), has been used extensively to predict the accumulation of laser damage from photothermal inactivation of critical cellular proteins. Damage from photochemical processes is less well studied and most models have not been verified because they require quantification of one or more uncharacterized chemical species. Additionally, few reports on photochemical damage report temperature history, measured or simulated. We used simulated threshold temperatures from a previous in vitro study to distinguish between photothermal and photochemical processes. Assuming purely photochemical processes also inactivate critical cellular proteins, we report the use of a photothermal Ω and a photochemical Ω that work in tandem to indicate overall damage accumulation. The combined damage integral (ΩCDI) applies a mathematical switch designed to describe photochemical damage relative to wavelength and rate of photon delivery. Although only tested in an in vitro model, this approach may transition to predict damage at the mammalian retina.

Copyright of Electric Power Automation Equipment / Dianli Zidonghua Shebei is the property of Electric Power Automation Equipment Press and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Image Salient Object Detection (SOD) is a fundamental research topic in the area of computer vision. Recently, the multimodal information in RGB, Depth (D), and Thermal (T) modalities has been proven to be beneficial to the SOD. However, existing methods are only designed for RGB-D or RGB-T SOD, which may limit the utilization in various modalities, or just finetuned on specific datasets, which may bring about extra computation overhead. These defects can hinder the practical deployment of SOD in real-world applications. In this paper, we propose an end-to-end U nified T riplet D ecoder Net work, dubbed UTDNet, for both RGB-T and RGB-D SOD tasks. The intractable challenges for the unified multimodal SOD are mainly two-fold, i.e., (1) accurately detecting and segmenting salient objects, and (2) preferably via a single network that fits both RGB-T and RGB-D SOD. First, to deal with the former challenge, we propose the multi-scale feature extraction unit to enrich the discriminative contextual information, and the efficient fusion module to explore cross-modality complementary information. Then, the multimodal features are fed to the triplet decoder, where the hierarchical deep supervision loss further enable the network to capture distinctive saliency cues. Second, as to the latter challenge, we propose a simple yet effective continual learning method to unify multimodal SOD. Concretely, we sequentially train multimodal SOD tasks by applying Elastic Weight Consolidation (EWC) regularization with the hierarchical loss function to avoid catastrophic forgetting without inducing more parameters. Critically, the triplet decoder separates task-specific and task-invariant information, making the network easily adaptable to multimodal SOD tasks. Extensive comparisons with 26 recently proposed RGB-T and RGB-D SOD methods demonstrate the superiority of the proposed UTDNet. [ABSTRACT FROM AUTHOR]

Named Entity Recognition (NER) can be divided into three subtasks: Flat, Nested, and Discontinuous. They are usually handled separately and independently, and most traditional approaches rely on representations in the form of Token and Span, which limits the flexibility of the models. Although Seq2Seq-based models have recently been proposed to handle these three NER tasks in a unified way, they usually focus too much on independent verification of entity spans while ignoring the label dependencies and detailed word-word relations in the modeling process. To this end, we propose the MLSU (Multi-level Semantic Understanding) model for unified NER tasks. The overall architecture leverages an Encoder-Decoder structure that integrates sequence and grid-level semantic blocks, enhancing the model’s ability to process complex text structures effectively. The model combines entity labels attention with pre-trained GloVe embeddings at the sequence level to optimize encoding and enhance context comprehension. Subsequently, the grid-level semantic extraction employs Conditional Layer Normalization(CLN) and Dilated CNNs. This enables MLSU to capture the interaction between text and labels effectively and enhances the ability to process complex text. We conducted experiments on six popular NER datasets and our MLSU model achieved generally good performance at or near the SOTA results, validating its effectiveness and advantages in handling various NER tasks.

Human-centric Computer Vision Tasks (HCTs) refer to a series of tasks related to the human body, such as Human Pose Estimation, Pedestrian Tracking, Re-Identification (ReID), Human Parsing and Action Recognition, etc. In the past three years, a large number of Human-centric Methods (HCMds) for HCTs have emerged, based on a common assumption that these tasks should share the same underlying semantic structure of the human body, integrating multi-modal information could achieve more powerful functionality or lower computational cost. However, a systematic and comprehensive literature review on this field is still missing, this survey provides a comprehensive review of these works. We first give a clear definition and taxonomy standard, then propose a new taxonomy of HCMds. Next, we discuss key technologies and the path towards Unified Human-centric model (UniHCM). Third, following the new taxonomy, we take a brief review and summary highlights of the representative model, framework, and benchmark for HCTs. Finally, We discuss and analyze the limitations of existing HCMds and suggest possible future research directions.

The low adoption rate and lack of industry-level large-scale applications are prominent problems facing the application of blockchain technology in the food supply chain field today. By integrating existing fragmented research, this paper provides a more systematic model of enterprise blockchain technology adoption behavior, which can not only provide theoretical support for academia to explore and understand enterprise blockchain technology adoption behavior, but also provide references for enterprises to promote blockchain project implementation decisions. By reviewing and combing relevant studies both domestic and overseas, integrating the core viewpoints of behavior reasoning theory, technology-organization-environment framework, innovation diffusion, and social capital, this paper constructs an unified model of enterprise blockchain technology adoption behavior in food supply chain field, and theoretically and normatively explains the antecedent factors that affect enterprise blockchain technology adoption intention. The model shows that enterprises form technology adoption intention by reasoning and judging 23 push and constraint factors under three dimensions of technological characteristics, organizational characteristics, and environmental characteristics, which drive technology adoption behavior. Therefore, enterprises should focus on strengthening the impact of push factors and preventing the impact of constraint factors to promote the adoption and implementation of blockchain technology more effectively.

When the near-field and far-field information of a target is uncertain, it is necessary to choose a suitable localization method. The modified polar representation (MPR) method integrates the two scenarios and achieves a unified localization with direction of arrival (DOA) estimation in the far field and position estimation in the near field. Previous studies have only proposed solutions for stationary environments and have not considered the motion factor. Therefore, this paper proposes a new unified positioning algorithm using multi-sensor time difference of arrival (TDOA) and frequency difference of arrival (FDOA) measurements without prior target source information. The method represents the position of the target source using MPR and describes the localization problem as a weighted least squares (WLS) problem with two constraints. We first obtain the initial estimates by WLS without considering the constraints and then investigate a two-step error correction method based on the constraints. The first step corrects the initial estimate using the Taylor series expansion technique, and the second step corrects the DOA estimate in the previous step using the direct error compensation technique based on the properties of the second constraint. Simulation experiments show that the method is effective for the unified positioning of moving targets and can achieve the Cramer–Rao lower bound (CRLB).

This study introduces a unified two-way (punching) shear design model for concrete reinforced with slabs steel- and FRP bars. The unified model modifies the ACI 318-19 punching shear model to account for the axial stiffness of the tension reinforcement by providing a new modification term (ρ Ereft./Ec)0.26, representing the reinforcing ratio multiplied by the elastic modular ratio of longitudinal bars to concrete. The unified punching shear model has been evaluated over two experimental datasets: FRP-RC slabs (145 specimens) and steel-RC slabs (390 specimens). A comparative assessment between the ACI 440.11-22, CSA S806, JSCE, and the proposed models to test their generalization abilities for slabs reinforced with both types of reinforcement. The assessment indicated a similar performance of the proposed model to the CSA S806 and the JSCE models in terms of statistical measures. The ACI 440.11–22 showed a high average strength ratio (Vexp./Vpred.) of 1.83. However, the ACI and the CSA models showed reduction in the conservatism at high reinforcement axial stiffness. In addition, reliability analysis using Monte Carlo simulation indicated that the unified punching shear model provides a satisfactory safety level with a reliability index between 3.5 and 4.0 for both steel and FRP-RC slabs.

In this work, a unified thermodynamic model for isothermal feedstock gasification is proposed to predict the gasification performance at carbon boundary point (CBP), by considering total hydrogen/total carbon, total oxygen/total carbon (entering into the system) and gasification temperature as unifying variables. The unified model is applicable for any combination of feedstocks, any blend of gasification agents O 2 , steam and CO 2 and any practically relevant gasification temperature. The simple model can be solved analytically to predict the amount of unknown gasification agent required to reach CBP. The unified model is also used to predict syngas composition, cold gas efficiency and amount of heat to be supplied, at CBP. The model is validated for different feedstocks, blend of gasification agents and gasification temperatures. The results predicted by the model are within 10% of the results from Aspen Plus. [Display omitted] • Unified model for optimal performance prediction of isothermal gasifier is proposed. • Three variables unifying the optimal performance are identified. • Unified model is solved using graphical and analytical approaches. • Unified model applicable for combination of feedstocks/gasification agents. • Model predictions using analytical approach are within 10% of Aspen Plus results. [ABSTRACT FROM AUTHOR]

Determination of fiber-reinforced polymer (FRP) confinement-induced improvements in the mechanical properties of concrete columns under compression is a current concern, particularly if partial confinement applied on a noncircular cross-sectional shape is to be considered. Although several design-oriented predictive formulations have been proposed for the calculation of axial strength and axial strain ductility of FRP-confined concrete, their applications are, in general, limited to a specific cross-sectional shape (circular, square, or rectangular cross section) and a certain confinement arrangement (fully or partially confining system). Accordingly, the aim in this study is to establish new unified strength and ductility models for concrete columns of circular or noncircular cross sections with fully or partially confining FRP systems. To achieve the highest level of predictive performance through a nonlinear regression technique, two datasets, consisting of 2,117 test data of peak strength and 2,050 test data of strain ductility, available in the literature, were collected. The dominance degrees of size effect, sectional noncircularity (corner radius ratio), cross-sectional aspect ratio, and confinement configuration type on confinement effectiveness were evaluated and reflected in the development of these regression-based models. Through predictions of test data compiled in the datasets and a comparison with the performances of available predictive models, the proposed unified formulations demonstrated a high level of reliability and were found to be proper for design purposes. [ABSTRACT FROM AUTHOR]

Copyright of Science & Technology of Food Industry is the property of Science & Technology of Food Industry Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Cinematic Shot Attribute Analysis aims to analyze the intrinsic attributes of movie shots, such as movement and scale. In previous methods, specialized architectures were designed for each specific task and relied on the use of optical flow maps. In this paper, we consider shot attribute analysis as a unified task of motion–static weight allocation, and propose a motion–static dual-path architecture for recognizing various shot attributes. In this architecture, we design a new action cue generation module for adapting the end-to-end training process instead of a pre-trained optical flow network; and, to address the issue of limited samples in movie shot datasets, we design a fixed-size adjustment strategy to enable the network to directly utilize pre-trained vision transformer models while adapting to shot data inputs at arbitrary sample rates. In addition, we quantitatively analyze the sensitivity of different shot attributes to motion and static features for the first time. Subsequent experimental results on two datasets, MovieShots and AVE, demonstrate that our proposed method outperforms all previous approaches without increasing computational cost. [ABSTRACT FROM AUTHOR]

A unified numerical model for simulating thermal flows past and through porous media whose thermal properties are identical with those of the surrounding fluid is proposed in this paper, which advances the work of the unified immersed boundary‐lattice Boltzmann flux solver (UIB‐LBFS). The proposed method greatly simplifies the conventional computation procedure for thermal flows past porous bodies by unifying the governing equations for the flow and temperature fields in both the porous and fluid domains. This is achieved by introducing a diffuse layer, across which the variation of the flow properties including its thermal features are expected to be smooth. Also, the developed method can be applied on either uniform or nonuniform grid since the numerical fluxes are locally reconstructed by LBFS at cell interfaces. In addition, the boundary conditions for both the flow and temperature fields at the porous‐fluid interface with complex shape can be implemented through IBM in a simple way. The proposed method is validated by several numerical examples over wide ranges of thermal parameters from forced convection to mixed convection. Excellent agreements are obtained for all simulations, including the one with severe temperature gradient in the thin temperature boundary layer at Pr=10$$ \mathit{\Pr}=10 $$, which demonstrates the capabilities of the proposed method. [ABSTRACT FROM AUTHOR]

In this work, within the framework of a unified model for the discharge gap and electrodes, a comparative numerical analysis was carried out on the effect of evaporation of graphite anode material on the characteristics of the arc discharge in helium and argon. The effect of changing the plasma-forming ion, in which the ion of evaporated atomic carbon becomes the dominant ion, is demonstrated. For an arc discharge in helium, this effect is accompanied by a jump-like change in the dependence of the current density on voltage (CVC), and smoothly for a discharge in argon. With regard to the dynamics of the ignition of an arc discharge, it is shown that during the transition from glow discharge to arc in helium, the discharge parameters are also accompanied by an abrupt change, while in argon, this transition is smooth. This is due to the fact that the ionization potentials, as well as the ionization cross sections, differ significantly for helium and carbon, and are close in value for helium and argon. For various points on the CVC, the density distributions of the charged and neutral particles of an inert gas and evaporated gases are presented. [ABSTRACT FROM AUTHOR]

Traditional systems typically require different models for processing different modalities, such as one model for RGB images and another for depth images. Recent research has demonstrated that a single model for one modality can be adapted for another using cross-modality transfer learning. In this paper, we extend this approach by combining cross/inter-modality transfer learning with a vision transformer to develop a unified detector that achieves superior performance across diverse modalities. Our research envisions an application scenario for robotics, where the unified system seamlessly switches between RGB cameras and depth sensors in varying lighting conditions. Importantly, the system requires no model architecture or weight updates to enable this smooth transition. Specifically, the system uses a depth sensor in low light conditions (night time) and both an RGB camera and a depth sensor or RGB camera only in well-lit environments. We evaluate our unified model on the SUN RGB-D dataset and demonstrate that it achieves a similar or better performance in terms of the mAP50 compared to state-of-the-art methods in the SUNRGBD16 category and a comparable performance in point-cloud-only mode. We also introduce a novel inter-modality mixing method that enables our model to achieve significantly better results than previous methods. We provide our code, including training/inference logs and model checkpoints, to facilitate reproducibility and further research. [ABSTRACT FROM AUTHOR]

Background: The degree to which suicide risk aggregates in US families is unknown. The authors aimed to determine the familial risk of suicide in Utah, and tested whether familial risk varies based on the characteristics of the suicides and their relatives. Methods: A population-based sample of 12 160 suicides from 1904 to 2014 were identified from the Utah Population Database and matched 1:5 to controls based on sex and age using at-risk sampling. All first through third- and fifth-degree relatives of suicide probands and controls were identified (N = 13 480 122). The familial risk of suicide was estimated based on hazard ratios (HR) from an unsupervised Cox regression model in a unified framework. Moderation by sex of the proband or relative and age of the proband at time of suicide (<25 v. ⩾25 years) was examined. Results: Significantly elevated HRs were observed in first- (HR 3.45; 95% CI 3.12–3.82) through fifth-degree relatives (HR 1.07; 95% CI 1.02–1.12) of suicide probands. Among first-degree relatives of female suicide probands, the HR of suicide was 6.99 (95% CI 3.99–12.25) in mothers, 6.39 in sisters (95% CI 3.78–10.82), and 5.65 (95% CI 3.38–9.44) in daughters. The HR in first-degree relatives of suicide probands under 25 years at death was 4.29 (95% CI 3.49–5.26). Conclusions: Elevated familial suicide risk in relatives of female and younger suicide probands suggests that there are unique risk groups to which prevention efforts should be directed – namely suicidal young adults and women with a strong family history of suicide. [ABSTRACT FROM AUTHOR]

Despite the many axial confinement models already proposed for the determination of the peak compressive strength of fiber‐reinforced polymer (FRP) confined concrete columns, they are, in general, applicable only to concrete columns of circular or square cross‐section, with full or partial confinement arrangements. In this study, by proposing a cross‐sectional and confining system unification approach, a new model is developed and calibrated based on a large test database. For the generalization of the cross‐section and FRP‐based confinement arrangement, the concept of confinement efficiency factor with a unified mathematical framework is adopted. By simulating experimental tests and comparing to the predictions of existing confinement models, the developed one demonstrates a very high reliability and suitable for design purposes by balancing the simplicity of the usage and accuracy. [ABSTRACT FROM AUTHOR]

Carbon fiber-reinforced plastic (CFRP) bonded, riveted, and adhesive-rivet hybrid joints are among the weak parts of a structure. To study their strength characteristics and strength characterization methods, the key parameters of the CFRP constitutive model were obtained by constitutive tests, based on which finite element (FE) models of the three types of CFRP joints were established. The failure modes and F – D (load-displacement) curves of the CFRP joints were obtained through quasi-static tensile tests, and the accuracy of the FE models was verified. Based on the test results, the test strength models of the three types of CFRP joints were established, along with a unified model to characterize the joint strength; however, the strength models were affected by the joint parameters. Therefore, the influence of nine parameters on the joint strength were discussed, and the unified strength models of the three types of CFRP joints under each influencing parameter were established based on the initial strength model. The results showed that the unified strength models were highly correlated with R2 > 0.92. These results provide useful references for the strength check and parameter optimization of CFRP joints. [Display omitted] ● Constitutive models and failure modes of three kinds of CFRP joints were obtained. ● The test strength models were established by analyzing the strength characteristics of the three joints. ● The influence of nine parameters on the strength of the three joints was analyzed. ● The unified strength models of the three joints under each influencing parameter were established. [ABSTRACT FROM AUTHOR]

Data in real world are usually characterized in multiple views, including different types of features or different modalities. Multi-view learning has been popular in the past decades and achieved significant improvements. In this paper, we investigate three challenging problems in the field of incomplete multi-view representation learning, namely, i) how to reduce the influences produced by missing views in multi-view dataset, ii) how to learn a consistent and informative representation among different views and iii) how to alleviate the impacts of the inherent noise in multi-view data caused by high-dimensional features or varied quality for different data points. To address these challenges, we integrate these three tasks into a problem and propose a novel framework termed Noise-aware Incomplete Multi-view Learning Networks (NIM-Nets). NIM-Nets fully utilize incomplete data from different views to produce a multi-view shared representation which is consistent, informative and robust to noise. We model the inherent noise in data by defining the distribution $\Gamma $ and assuming that each observation in the incomplete dataset is sampled from the distribution $\Gamma $. To the best of our knowledge, this is the first work to unify learning the consistent and informative representation, alleviating the impacts of noise in data and handling the view-missing patterns in multi-view learning into a framework. We also first give a definition of robustness and completeness for incomplete multi-view representation learning. Based on NIM-Nets, we present joint optimization models for classification and clustering, respectively. Extensive experiments on different datasets demonstrate the effectiveness of our method over the existing work based on classification and clustering tasks in terms of different metrics. [ABSTRACT FROM AUTHOR]

Calibration is an important premise for vision-guided robot system, especially for precise industrial operations on curved surfaces. Most of existing methods focus on one or partial union of errors including assembly error of the robot, relative pose error between the vision unit and the robot, or installation error of end effector, which is still challenging in practice. To solve this problem, this article proposes a new calibration method based on error correction matrix by which different manifestations of each error is unified. As the nonlinear equations formed by these matrices cannot be solved analytically, a numerical optimization solution based on Lie algebra is presented. Specifically, the matrices acquisition is formulated as the problem of minimizing the sum of distance deviations between actual and ideal tool center points. This problem is then solved by differentiating these matrices with a form of left multiplicative perturbation. In this way, a high precision joint calibration with multisource errors is achieved. The proposed method is verified by simulations and experiments. [ABSTRACT FROM AUTHOR]

Many m-gov adoption models have been proposed, which can confound researchers and policymakers. In this article, we reviewed 17 studies on m-gov adoption, identifying 25 different factors. We conducted two focus groups to discuss the adequacy of these factors, generating the first version of a unified model. We tested this model using data from 806 survey respondents from Brazil. The proposed unified model is parsimonious and outperforms other theoretical models. [ABSTRACT FROM AUTHOR]

The dowel effect in hybrid bonded fiber reinforced plastic (HB-FRP) to concrete joints is difficult to quantify and make the anchoring joint instability. A new HB-FRP joint was built, which eliminates bonding between steel and FRP plates. Based on an empirical model of external bonding FRP (EB-FRP) to concrete joints and the new model of slip at debonding, the new unified bond-slip model of the HB-FRP joints is proposed. The parameters that affect the bond-slip model are analyzed. Pull-off tests are conducted for eight specimens with different numbers of mechanical fasteners, which has verified the proposed model. The load-slip relationship at the loaded end of the FRP plate was well predicted by the partial interaction numerical calculation and further verify the theoretical model. [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel unsupervised continual-learning generative adversarial network for unified image fusion, termed as UIFGAN. In our model, for multiple image fusion tasks, a generative adversarial network for training a single model with memory in a continual-learning manner is proposed, rather than training an individual model for each fusion task or jointly training multiple tasks. We use elastic weight consolidation to avoid forgetting what has been learned from previous tasks when training multiple tasks sequentially. In each task, the generation of the fused image comes from the adversarial learning between a generator and a discriminator. Meanwhile, a max-gradient loss function is adopted for forcing the fused image to obtain richer texture details of the corresponding regions in two source images, which applies to most typical image fusion tasks. Extensive experiments on multi-exposure, multi-modal and multi-focus image fusion tasks demonstrate the advantages of our method over the state-of-the-art approaches. • We propose an unsupervised continual-learning GAN for training a single model. • The continual-learning GAN is proposed for unified image fusion tasks for the first time. • The discriminator distinguishes fused image from two source images simultaneously. • We adopt a max-gradient function for unified image fusion tasks. • The qualitative and quantitative results are superior to the state-of-the-art. [ABSTRACT FROM AUTHOR]

In order to solve the information overload problem on social media platforms and help users quickly capture the required information, in this study the problem of hashtag recommendation based on multimodal content was investigated. To address the heterogeneous differences between different modalities, a co-attention mechanism was used to model and fuse features of cross-modal content, and use Seq2Seq framework was used to generate new hashtag sequences to address the deficiency that multi-label classification methods could only recommend hashtags in the hashtag space of the dataset. An aggregation strategy was used to aggregate the recommendation results of classification methods into the generated hashtag sequences to obtain a unified recommendation model for both methods. The experimental results on a large-scale dataset showed that, firstly, the multimodal approach was more advantageous than the unimodal approach, and the unified recommendation model proposed in this paper had 9. 44 percentage points improvement in F1 value over the comparison model using unimodal approach, and 3.41 percentage points improvement over the comparison model using the classification method. Finally, the unified recommendation model UNIFIED-CO-ATT is 1.25 percentage points higher than GEN-CO-ATT in F1 values. The model proposed in this study could combine the advantages of classification and generation methods and could make the recommended hashtags have the advantages of accuracy and novelty at the same time. [ABSTRACT FROM AUTHOR]