Bidirectional Mapping Research Articles

Estimating Addiction-Related Brain Connectivity by Prior-Embedding Graph Generative Adversarial Networks.

The study of nicotine addiction mechanism is of great significance in both nicotine withdrawal and brain science. The detection of addiction-related brain connectivity using functional magnetic resonance imaging (fMRI) is a critical step in study of this mechanism. However, it is challenging to accurately estimate addiction-related brain connectivity due to the low-signal-to-noise ratio of fMRI and the issue of small sample size. In this work, a prior-embedding graph generative adversarial network (PG-GAN) is proposed to capture addiction-related brain connectivity accurately. By designing a dual-generator-based scheme, the addiction-related connectivity generator is employed to learn the feature map of addiction connection, while the reconstruction generator is used for sample reconstruction. Moreover, a bidirectional mapping mechanism is designed to maintain the consistency of sample distribution in the latent space so that addiction-related brain connectivity can be estimated more accurately. The proposed model utilizes prior knowledge embeddings to reduce the search space so that the model can better understand the latent distribution for the issue of small sample size. Experimental results demonstrate the effectiveness of the proposed PG-GAN.

The bidirectional mapping of colour metaphor in power: The effect of colour-semantic integration and processing depth.

Previous studies found that the abstract concept of power is embodied within space and weight, but there is not a consistent conclusion on the metaphorical relationship between power and colour. The present study adopted three experiments to investigate the metaphorical association between the concept of power and colour perception in Chinese, and how semantic processing and colour perception influence this mapping. Experiment 1 studied the effect of colour perception on the processing of power. The results showed that there was a faster response when high-power words were presented in gold and low-power words were presented in grey. Experiment 2 explored whether priming semantics of power affected colour judgement, and found no significant difference between gold and grey, neither in high-power words nor low-power words. Experiment 3 discovered that participants preferred to connect the golden pseudo words with high-power words and connect the grey pseudo words with low-power words. Overall, the present study substantiated a two-way metaphorical link between colour and power. A metaphorical association is also influenced by the degree of integration of colour into semantics and the depth of semantic processing.

Enhancing low-light images via dehazing principles: Essence and method

Given the visual resemblance between inverted low-light and hazy images, dehazing principles are borrowed to enhance low-light images. However, the essence of such methods remains unclear, and they are susceptible to over-enhancement. Regarding the above issues, in this letter, we present corresponding solutions. Specifically, we point out that the Haze Formation Model (HFM) used for image dehazing exhibits a Bidirectional Mapping Property (BMP), enabling adjustment of image brightness and contrast. Building upon this property, we give a comprehensive and in-depth theoretical explanation for why dehazing on inverted low-light image is a solution to the image brightness enhancement problem. Further, an Adaptive Full Dynamic Range Mapping (AFDRM) method is then proposed to guide HFM in restoring the visibility of low-light images without inversion, while overcoming the issue of over-enhancement. Extensive experiments validate our proof and demonstrate the efficacy of our method.

Improved MTPA and MTPV Optimal Criteria Analysis Based on IPMSM Nonlinear Flux-Linkage Model

The use of interior permanent-magnet synchronous machines (IPMSMs) is prevalent in automotive and vehicle traction applications due to their high efficiency over a wide speed range. Given the high-power-density requirements of automotive IPMSMs, it is imperative to consider the effect of nonlinearities, such as saturation and cross-coupling, on the motor model. The aforementioned nonlinearities render conventional linear motor models incapable of accurately describing the operating characteristics of the IPMSM, including the maximum torque per ampere (MTPA) trajectory, the flux-weakening (FW) trajectory, and the maximum torque per volt (MTPV) trajectory. With respect to the linear motor model, the nonlinear flux-linkage model is gradually receiving attention from researchers. This modeling method represents the nonlinear behavior of the motor through the direct establishment of a bidirectional mapping relationship between flux-linkage and current. It is capable of naturally incorporating the effects of magnetic saturation and cross-coupling factors. However, the analysis of the current trajectory optimal criteria based on this model has not yet been reported. In this paper, the optimal criteria for the MTPA and MTPV current trajectories are analyzed based on the nonlinear flux-linkage model of IPMSMs. Firstly, the nonlinear flux-linkage model of the tested IPMSM is established by the experimental calibration method. The mathematical analytical expressions of the MTPA and MTPV optimal criteria are then analyzed by constructing and solving optimal problems with different objectives. Finally, the current command table applicable to actual motor control is constructed by calculating the current command for different operating conditions according to the optimal criteria proposed in this paper. The validity and feasibility of the optimal criteria proposed in this paper are verified through experimental tests on different operating conditions.

EduCross: Dual adversarial bipartite hypergraph learning for cross-modal retrieval in multimodal educational slides

In the digital education landscape, cross-modal retrieval (CMR) from multimodal educational slides represents a significant challenge, particularly because of the complex nature of academic content, which includes images, diagrams, equations, and tables across various subjects such as mathematics and biology. Current CMR systems are primarily designed for “(natural) image to text” interactions (or vice versa) and inadequately address real-world educational scenarios. This study presents EduCross, a novel framework devised to enhance CMR within multimodal educational slides, which is a domain in which traditional retrieval systems fall short. Recognizing the imperative for a system that is tailored to the educational context, EduCross integrates dual adversarial bipartite hypergraph learning, harnessing the capabilities of generative adversarial networks with figure-text dual channels. This powerful combination facilitates robust bidirectional mapping, allowing for the precise association of figures with their descriptive spoken language segments and ensuring a comprehensive CMR experience. Specifically, we develop framelet-based deep bipartite hypergraph neural networks that effectively manage the high-order relationships between diverse educational content types and various types of slide figures. Our experimental results underscore the superior performance of EduCross, demonstrating its effectiveness through the use of the real Multimodal Lecture Presentations dataset that mirrors authentic educational settings. These outcomes highlight the significant advancements of EduCross over existing methods, marking a leap forward in the accurate retrieval of multimodal educational content.

Bidirectional mapping between rhombic triacontahedron and icosahedral hexagonal discrete global grid systems

ABSTRACT The icosahedron is currently the mainstream polygon in research and application of discrete global grid systems (DGGS). However, compared to the rhombic triacontahedron (RT), the icosahedron has disadvantages, such as lower sphere-fitting accuracy, greater projection distortion, and difficulty in incorporating the matrix structure for geospatial data storage. More importantly, the special positional relationship between the rhombic triacontahedron and the Earth enables it to effectively support event simulations related to geographical locations. To this end, bidirectional mapping of the hexagonal grid between the RT and icosahedron was proposed, which can efficiently integrate the existing datasets and algorithms of icosahedral DGGS into RT DGGS, thereby achieving seamless conversion between heterogeneous grid systems. We established geometric and topological correlations between the RT and icosahedron, abstracted the spatial algebraic structures of hexagonal grids on the two different polygons, and constructed mapping relationships between them. Finally, conversion between heterogeneous grid indices was achieved using dual quaternions. Experiments revealed that the proposed method was 3.9150 and 2.8151 times more efficient at grid conversion from RT to icosahedron and from icosahedron to RT, respectively, than was a method using latitude/longitude coordinates as a medium.

Footstone of Metaverse: A Timely and Secure Crowdsensing

Recently, the metaverse has been a hot research topic that fuses multiple cutting-edge techniques, including virtual reality (VR), augmented reality (AR), artificial intelligence (AI), Internet of Things (IoT), and 5th generation (5G) networks. Supported by these powerful cornerstones, people can establish a hyper spatiotemporal virtual world that seamlessly links to the real world and create personal activities such as art creation, concert display, and stock trading. Virtual-reality mapping demands enormous information transmission between two worlds without sensible latency since delayed data could severely impact user experiences. Besides, intrinsic flaws in fundamental technologies also aggravate the vulnerabilities in the metaverse due to real-time interaction. Hence, this paper first proposes a timely and secure data collection framework based on crowdsensing for metaverse modeling. In our proposed Data encryption, Transmission and Perception optimization (DTP) model, the metaverse servers can recruit workers to gather environmental parameters and user information through five layers architecture (i.e., perception layer, link layer, transmission layer, operation layer, and application layer). Each layer takes responsibility for data computation and encryption, which can stimulate data dissemination in a fast and reliable manner for the metaverse. Furthermore, we present a healthcare use case that introduces a bidirectional mapping of patient data between physical and virtual space. This use case illustrates the prospects of the DTP model. Finally, we conclude this paper and summarize several significant future research directions.

Development of Machine Tool Disassembly and Assembly Training and Digital Twin Model Building System based on Virtual Simulation

In response to the issues of poor safety and insufficient training equipment in traditional machine tool assembly and disassembly training, this paper proposes a machine tool assembly and disassembly training system based on virtual simulation. The system utilizes Unity 3D to construct an interactive virtual training environment, achieving highly simulated simulations and teaching of the machine tool assembly and disassembly process. Additionally, the system is capable of establishing communication with actual machine tools, achieving bidirectional mapping between digital and physical spaces, allowing learners to observe the dynamic responses of machine tool models through the virtual environment. This virtual-real interaction mechanism enriches the application scenarios of the system, introducing digital twin technology into the virtual simulation of machine tool assembly and disassembly training. This paper also provides a detailed introduction to the design and implementation of core technologies such as the system's overall architecture, scene modeling, and virtual-real interaction, offering significant support for intelligent teaching in machine tool assembly and disassembly.

Reliable Event Generation With Invertible Conditional Normalizing Flow.

Event streams provide a novel paradigm to describe visual scenes by capturing intensity variations above specific thresholds along with various types of noise. Existing event generation methods usually rely on one-way mappings using hand-crafted parameters and noise rates, which may not adequately suit diverse scenarios and event cameras. To address this limitation, we propose a novel approach to learn a bidirectional mapping between the feature space of event streams and their inherent parameters, enabling the generation of reliable event streams with enhanced generalization capabilities. We first randomly generate a vast number of parameters and synthesize massive event streams using an event simulator. Subsequently, an event-based normalizing flow network is proposed to learn the invertible mapping between the representation of a synthetic event stream and its parameters. The invertible mapping is implemented by incorporating an intensity-guided conditional affine simulation mechanism, facilitating better alignment between event features and parameter spaces. Additionally, we impose constraints on event sparsity, edge distribution, and noise distribution through novel event losses, further emphasizing event priors in the bidirectional mapping. Our framework surpasses state-of-the-art methods in video reconstruction, optical flow estimation, and parameter estimation tasks on synthetic and real-world datasets, exhibiting excellent generalization across diverse scenes and cameras.

A Normalizing Flow-Based Bidirectional Mapping Residual Network for Unsupervised Defect Detection

A Normalizing Flow-Based Bidirectional Mapping Residual Network for Unsupervised Defect Detection

Bidirectional visual-tactile cross-modal generation using latent feature space flow model

Inspired by visual-tactile cross-modal bidirectional mapping of the human brain, this paper introduces a novel approach to bidirectional mapping between visual and tactile data, an area not fully explored in the predominantly unidirectional existing studies. First, we adopt separate Variational AutoEncoder (VAE) models for visual and tactile data. Furthermore, we introduce a conditional flow model built on the VAE latent feature space, enabling cross-modal bidirectional mapping between visual and tactile data using one model. The experimental results show that our method achieves excellent performance in terms of the similarity between the generated data and the original data (Structural Similarity Index (SSIM) of visual data: 0.58, SSIM of tactile data: 0.80), the classification accuracy on generated data (visual data: 91.60%, tactile data: 88.05%), and the zero-shot classification accuracy between generated data and language (visual data: 44.49%, tactile data: 45.03%). To the best of our knowledge, the method proposed in this paper is the first one to utilize a single model to achieve bidirectional mapping between visual and tactile data. Our model and code will be made public after the acceptance of the paper.

A Comprehensive Model Incorporating Multiple Spatial Relations in 3D Space

Aims: A Comprehensive Model Incorporating Multiple Spatial Relations in 3D Space. Background: At present, the research on two-dimensional spatial relation expression and inference models is relatively mature, but these models cannot be used to deal with three-dimensional spatial orientation relations. With the application of spatial orientation relations, threedimensional spatial orientation relations are involved in many fields such as urban architectural design, robotics, image processing, etc. Two-dimensional spatial orientation relations models cannot satisfy the needs of three-dimensional spatial applications, so there is an urgent need to research three-dimensional spatial object orientation relations expression and inference models. Objective: This work aims to provide a comprehensive model incorporating multiple spatial relations in 3D space. The joint representation of direction, distance, and topological relations describes more complex spatial position relations among objects. Methods: Based on this comprehensive model, the computational properties of interval algebra are used to combine the directional and topological relations. Result: The study lays a good foundation for the formal representation and reasoning of spatial relations between regions, enhances the analyzability of spatial relations between objects, and improves the accuracy of spatial analysis. Conclusion: The main novel contribution of this paper is that we propose a comprehensive orientation relation model, called 3D-TRD, which considers three spatial orientation relations simultaneously. The paper gives examples to represent the position relations of two spatial objects by comparing the RCC8 model, the 3DR46 model, and the comprehensive model to highlight the advantages of our proposed model. Based on the model, the bidirectional mapping association method is also used to represent the location of the spatial objects. The first advantage of the 3DTRD model is that it represents spatial location relations more accurately than 3DR46, RCC8, and five qualitative distances. The second advantage of the 3D-TRD model is that it proposes a bidirectional mapping representation in three-dimensional space. The third advantage of the 3DTRD model is that it provides a good basis for the formal representation and inference study of the spatial relations between regions.

A multigrid partition coupled Eulerian–Lagrangian method for fluid–solid interaction problems

In this paper, a multigrid partition coupled Eulerian–Lagrangian method (PCELM) is proposed to model fluid–solid interactions between fluids and moving or fixed rigid bodies. In this method, Lagrangian particles are arranged in Eulerian grids to track the interface, and the partition weighted bidirectional mapping method is used to improve the mapping accuracy between particles and grids. The multigrid domain is established to realize independent and contact calculations of the fluid and solid, which can solve for the cross-interference of multiple velocity fields in a single grid domain. Moreover, based on the contact in the particle volume domain, a more accurate fluid–solid interaction method, in which the contact properties of real grid nodes are identified by judgment and search criteria and then corrected, is proposed to overcome false contact and nonphysical penetration problems. Then, the accuracy and stability of the PCELM are verified using representative fluid–solid interactions. Finally, experiments and simulations of dam-break floods impacting the dam board are carried out. These experiments prove that the PCELM can accurately capture the free surface of dam-break floods and can provide suggestions for the installation of dam boards in hydraulic engineering.

Bidirectional Mapping Perception-enhanced Cycle-consistent Generative Adversarial Network for Super-resolution of Brain MRI images.

As an effective tool for visualizing neurodegeneration, high-resolution structural magnetism facilitates quantitative image analysis and clinical applications. Super-resolution reconstruction technology allows to improve the resolution of images without upgrading the scanning hardware. However, existing super-resolution techniques relied on paired image data sets and lacked further quantitative analysis of the generated images. In this study, we proposed a semi-supervised generative adversarial network (GAN) model for super-resolution of brain MRI, and the synthetic images were evaluated using various quantitative measures. This model adopted the cycle-consistency structure to allow for a mixture of unpaired data for training. Perceptual loss was further introduced into the model to preserve detailed texture features at high frequencies. 363 subjects with both high-resolution (HR) and low-resolution (LR) scans and 217 subjects with HR scans only were used for model derivation, training, and validation. We extracted multiple voxel-based and surface-based morphological features of the synthetic and real 3D HR images for comparison. We further evaluated the synthetic images in the differential diagnosis of diseases. Our model achieved superior mean absolute error (0.049±0.021), mean squared error (0.0059±0.0043), peak signal-to-noise ratio (29.41±3.71), structural similarity index measure (0.914±0.048). Eight morphological metrics, both voxel-based and surface-based, showed significant agreement (P<0.0001). The gap of accuracy in disease diagnosis between synthetic and real HR images was within 5% and significantly outperformed the LR images. Our proposed model enables the reconstruction of HR MRI and could be used accurately for image quantification.Clinical relevance- Quantitative evaluation of the synthetic high-resolution images was used to determine whether the synthetic images have sufficient realism and diversity.

BMISP: Bidirectional mapping of image signal processing pipeline

After being processed by the image signal processing (ISP) pipeline in digital cameras, the sRGB images are nonlinear, and thus are not suitable for the computer vision tasks which work best in a linear color space. Therefore, mapping nonlinear sRGB images back to a linear color space is a highly valuable task. To achieve an accurate mapping, this paper proposes a framework based on convolutional neural networks, which models the ISP pipeline in both reverse and forward directions. In particular, for the reverse mapping, a U-net structure is applied to extract features from a given sRGB image, and the extracted features are utilized to estimate the linear and nonlinear transformations in the ISP pipeline. For the forward mapping, the original sRGB image is used as a guidance to embed the camera-style information. To incorporate the encoded prior knowledge, affine transformations are employed to modulate the features. Experiments demonstrate that the proposed framework is able to achieve the state-of-the-art performance.

The deep-learning-based evolutionary framework trained by high-throughput molecular dynamics simulations for composite microstructure design

The interfacial shear behavior, including the elastic deformation stage, the debonding stage and the sliding stage, is of critical importance in determining the overall performance of a composite. Here a deep-learning-based framework that integrates the generative adversarial network (GAN) and the genetic algorithm (GA) is proposed to establish a bidirectional mapping between the interfacial microstructure and the interfacial shear behavior of SiC fiber reinforced SiC composite (SiCf/SiC). High-throughput molecular dynamics (MD) simulations are carried out to collect the strain–stress curves of the SiCf/SiC interfaces under shear tractions, considering the exhaustive variations in the interfacial microstructure and the temperature. The well-trained GAN model can predict the shear strain–stress curve of an arbitrarily given interfacial microstructure with excellent robustness and accuracy. In the reverse direction, the GA is utilized to infer the interfacial microstructure given desired shear properties. We demonstrate that the proposed GAN-GA protocol can enhance the physical modeling of conventional continuum-scale simulation and efficiently identify the microstructure leading to optimal mechanical properties. Such capability can greatly facilitate the demand-oriented development of composite materials, thereby contributing to intelligent manufacturing.

Petri Nets-Based Modeling Solution for Cyber–Physical Product Control Considering Scheduling, Deployment, and Data-Driven Monitoring

For a complex electromechanical product that is a cyber–physical system (CPS), its dynamic behaviors are embodied in the closed-loop control between the logic process in its cyber component and actual actuators/sensors in its physical component, and thus, a well-defined model of the control is important to create a digital twin that acts as much like the real machine as possible. This article proposes a Petri nets (PNs)-based modeling solution that employs hybrid PNs (HPNs) for physics and system of sequential systems with shared resources (S4R) nets for logic in building a hierarchical control model. We also present PNs technologies for implementing a smooth transition and bidirectional mapping from the virtual prototype to the real machine. These technologies involve a PNs integration of a reinforcement learning (RL) method for generating a workflow scheduling agent in design, an extension of PNs definitions that is compatible with the microcontroller for easy deployment in manufacturing, and an architecture of PNs execution recording for data-driven monitoring in service. A software kit is provided for the solution that includes an integrated development environment of PNs, tools for quickly building a virtual prototype, and a monitor server for remote data-driven monitoring. This solution is successfully applied in the development of a typical cyber–physical product case, namely, the chemiluminescence immunoassay (CLIA) analyzer.

Bidirectional Mapping with Contrastive Learning on Multimodal Neuroimaging Data.

The modeling of the interaction between brain structure and function using deep learning techniques has yielded remarkable success in identifying potential biomarkers for different clinical phenotypes and brain diseases. However, most existing studies focus on one-way mapping, either projecting brain function to brain structure or inversely. This type of unidirectional mapping approach is limited by the fact that it treats the mapping as a one-way task and neglects the intrinsic unity between these two modalities. Moreover, when dealing with the same biological brain, mapping from structure to function and from function to structure yields dissimilar outcomes, highlighting the likelihood of bias in one-way mapping. To address this issue, we propose a novel bidirectional mapping model, named Bidirectional Mapping with Contrastive Learning (BMCL), to reduce the bias between these two unidirectional mappings via ROI-level contrastive learning. We evaluate our framework on clinical phenotype and neurodegenerative disease predictions using two publicly available datasets (HCP and OASIS). Our results demonstrate the superiority of BMCL compared to several state-of-the-art methods.

One-Sided Unsupervised Image Dehazing Network Based on Feature Fusion and Multi-Scale Skip Connection

Haze and mist caused by air quality, weather, and other factors can reduce the clarity and contrast of images captured by cameras, which limits the applications of automatic driving, satellite remote sensing, traffic monitoring, etc. Therefore, the study of image dehazing is of great significance. Most existing unsupervised image-dehazing algorithms rely on a priori knowledge and simplified atmospheric scattering models, but the physical causes of haze in the real world are complex, resulting in inaccurate atmospheric scattering models that affect the dehazing effect. Unsupervised generative adversarial networks can be used for image-dehazing algorithm research; however, due to the information inequality between haze and haze-free images, the existing bi-directional mapping domain translation model often used in unsupervised generative adversarial networks is not suitable for image-dehazing tasks, and it also does not make good use of extracted features, which results in distortion, loss of image details, and poor retention of image features in the haze-free images. To address these problems, this paper proposes an end-to-end one-sided unsupervised image-dehazing network based on a generative adversarial network that directly learns the mapping between haze and haze-free images. The proposed feature-fusion module and multi-scale skip connection based on residual network consider the loss of feature information caused by convolution operation and the fusion of different scale features, and achieve adaptive fusion between low-level features and high-level features, to better preserve the features of the original image. Meanwhile, multiple loss functions are used to train the network, where the adversarial loss ensures that the network generates more realistic images and the contrastive loss ensures a meaningful one-sided mapping from the haze image to the haze-free image, resulting in haze-free images with good quantitative metrics and visual effects. The experiments demonstrate that, compared with existing dehazing algorithms, our method achieved better quantitative metrics and better visual effects on both synthetic haze image datasets and real-world haze image datasets.

A partition-coupled Eulerian–Lagrangian method for large-deformation simulation of compressible fluid

We present a partition-coupled Eulerian–Lagrangian method (PCELM) for accurately tracking a free interface and a contact discontinuity of the compressible fluid with large deformation. This method tracks the interface by arranging splittable Lagrangian particles on an Eulerian grid and adopts a partition-weighted bidirectional mapping between particles and grids using a cubic B-spline as interpolation function. PCELM suppresses oscillation of the discontinuous surface by this partition-weighted remapping method and solves the problem of numerical fracture by a particle splitting method. A virtual particle method is also proposed to deal with discontinuity of particle flow at the boundary and to maintain interpolation accuracy at the boundary. The conservation of mass, momentum, and energy of PCELM is proved by conservation analysis. Accuracy tests and simulations of discontinuous surfaces and free interfaces are performed to verify the accuracy and stability of PCELM. The results show that PCELM has strong energy conservation and low energy dissipation and that it is not only better at suppressing oscillations than the original method, but can also simulate a compressible fluid with large deformation more accurately than weighted essentially nonoscillatory schemes.