Alignment Scheme Research Articles

Multi-Position Inertial Alignment Method for Underground Pipelines Using Data Backtracking Based on Single-Axis FOG/MIMU.

The inertial measurement method of pipelines utilizes a Micro-Electro-Mechanical Systems Inertial Measurement Unit (MIMU) to get the three-dimensional trajectory of underground pipelines. The initial attitude is significant for the inertial measurement method of pipelines. The traditional method to obtain the initial attitude uses three-axis magnetometers to measure the Earth's magnetic field. However, the magnetic field in urban underground pipelines is intricate, which leads to the initial attitude being inaccurate. To overcome this challenge, a novel multi-position initial alignment method based on data backtracking for a single-axis FOG and a three-axis Micro-Electro-Mechanical Inertial Measurement Unit (MIMU) is proposed. Firstly, the configuration of the sensors is determined. Secondly, according to the three-point support structure of the pipeline measuring instrument, a three-position alignment scheme is designed. Additionally, an initial alignment algorithm using the data backtracking method is developed. In this algorithm, a rough initial alignment is conducted by the data from single-axis FOG, and a fine initial alignment is conducted by the data from FOG/MIMU. Finally, an experiment was conducted to validate this method. The experiment results indicate that the pitch and roll angle errors are less than 0.05°, and the azimuth angle errors are less than 0.2°. This improved the precision of the 3-D trajectory of underground pipelines.

AdaTreeFormer: Few shot domain adaptation for tree counting from a single high-resolution image

The process of estimating and counting tree density using only a single aerial or satellite image is a difficult task in the fields of photogrammetry and remote sensing. However, it plays a crucial role in the management of forests. The huge variety of trees in varied topography severely hinders tree counting models to perform well. The purpose of this paper is to propose a framework that is learnt from the source domain with sufficient labeled trees and is adapted to the target domain with only a limited number of labeled trees. Our method, termed as AdaTreeFormer, contains one shared encoder with a hierarchical feature extraction scheme to extract robust features from the source and target domains. It also consists of three subnets: two for extracting self-domain attention maps from source and target domains respectively and one for extracting cross-domain attention maps. For the latter, an attention-to-adapt mechanism is introduced to distill relevant information from different domains while generating tree density maps; a hierarchical cross-domain feature alignment scheme is proposed that progressively aligns the features from the source and target domains. We also adopt adversarial learning into the framework to further reduce the gap between source and target domains. Our AdaTreeFormer is evaluated on six designed domain adaptation tasks using three tree counting datasets, i.e. Jiangsu, Yosemite, and London. Experimental results show that AdaTreeFormer significantly surpasses the state of the art, e.g. in the cross domain from the Yosemite to Jiangsu dataset, it achieves a reduction of 15.9 points in terms of the absolute counting errors and an increase of 10.8% in the accuracy of the detected trees’ locations. The codes and datasets are available at https://github.com/HAAClassic/AdaTreeFormer.

Polarization detection for cylindrical vector beams empowered by pixelated metasurfaces

Polarimetry plays an indispensable role in the light–matter interactions. Nevertheless, conventional components developed for polarization measurements suffer from bulky volume and spatial alignment schemes, causing them to reveal limited performance in determining inhomogeneous polarization distributions. Here, we propose a polarization detection scheme based on pixelated all-dielectric metasurfaces using spin-multiplexing coding techniques. The polarization resolving capability of the pixelated metasurface under homogeneous linearly polarized illumination was first evaluated, and the extracted peak coordinates were used to establish an exact functional relationship with the azimuthal angle. Subsequently, the measurement of spatial inhomogeneous polarization was further explored with a focusing matrix assembled from pixelated metasurfaces. The proposed polarization detection strategy can be extended to other spectral bands without discrimination, stimulating potential applications in high-resolution imaging, sensing and data communication.

Facial Action Unit detection based on multi-task learning strategy for unlabeled facial images in the wild

Facial Action Unit (AU) detection often relies on highly-cost accurate labeling or inaccurate pseudo labeling techniques in recent years. How to introduce large amounts of unlabeled facial images in the wild into supervised AU detection frameworks has become a challenging problem. Additionally, nearly every type of AUs has the problem of unbalanced positive and negative samples. Inspired by other multi-task learning frameworks, we first propose a multi-task learning strategy boosting AU detection in the wild through jointing facial landmark detection and AU domain separation and reconstruction. Our introduced dual domains facial landmark detection framework can solve the lack of accurate facial landmark coordinates during the AU domain separation and reconstruction training process, while the parameters of homostructural facial extraction modules from these two similar facial tasks are shared. Moreover, we propose a pixel-level feature alignment scheme to maintain the consistency of features obtained from two separation and reconstruction processes. Furthermore, a weighted asymmetric loss is proposed to change the contribution of positive and negative samples of each type of AUs to model parameters updating. Experimental results on three widely used benchmarks demonstrate our superiority to most state-of-the-art methods for AU detection.

Heterogeneous heatmap distillation framework based on unbiased alignment for lightweight human pose estimation

The growing demand for mobile devices has generated interest in lightweight human pose estimation. Currently, lightweight estimation generally uses heatmap-based methods, which has demonstrated exceptional performance. However, their use of non-differentiable post-processing imposes considerable inference latencies. Conversely, integral-based approaches expedite the inference process by employing a soft-argmax operation but compromise in accuracy. Integrating explicit heatmap knowledge learned using the heatmap-based method into the implicit heatmap generated by the integral-based method, thereby combining the best of both worlds, offers a promising avenue. However, owing to the disparities in supervision and inference processes, the explicit and implicit heatmaps are heterogeneous. Consequently, direct transfer of knowledge presents difficulties in ensuring consistencies in heat value and location. In this paper, we propose a novel Heterogeneous Heatmap Distillation (HHD) framework that effectively tackles these challenges. The framework seamlessly integrates explicit heatmap knowledge that contains high-precision localization information into implicit heatmaps. The framework revolves around an unbiased heatmap alignment scheme encompassing two steps: heterogeneous heatmap normalization and unbiased cropping. Heterogeneous heatmap normalization separately normalizes the output feature maps of both the teacher and student models, alleviating potential heat value bias during the knowledge transfer. Unbiased cropping applies closed-form computation on the normalized teacher and student heatmap to eliminate location bias. Additionally, mirror expansion is implemented to handle potential cases wherein the cropped region extends beyond the image boundary. Extensive experiments demonstrate the efficiency and effectiveness of our methods on the MSCOCO and MPII datasets compared to other integral-based lightweight networks. Our source codes and pre-trained models are available at https://github.com/ducongju/HHD.

A rapid alignment method for high-precision rotational INS within group affine

In this paper, a rapid alignment method for high-precision rotation inertial navigation system (RINS) within group affine is proposed based on the Lie group theory. The proposed method can effectively solve the ultimate contradiction of initial alignment: speed and accuracy. It meets the requirements of fast response and high accuracy of high-precision RINS. We abandon the conventional coarse alignment + fine alignment scheme and realize a linear Kalman filter alignment method with arbitrarily initial misalignment based on the Earth-centered Earth-fixed frame RINS mechanization and the Lie group theory. Subsequently, considering the error characteristics of RINS two-position alignment, new left-invariant error model and observation model satisfying the group affine property are constructed to improve the convergence speed and accuracy of alignment. It is verified that the method can complete the initial alignment of arbitrary large misalignment and reach the ultimate accuracy in a short time, and the land vehicle navigation test shows that the proposed alignment method can complete the initial alignment of arbitrary misalignment in 5 min under the swaying disturbance and achieve better pure inertial navigation accuracy compared with the traditional scheme.

SyFormer: Structure-Guided Synergism Transformer for Large-Portion Image Inpainting

Image inpainting is in full bloom accompanied by the progress of convolutional neural networks (CNNs) and transformers, revolutionizing the practical management of abnormity disposal, image editing, etc. However, due to the ever-mounting image resolutions and missing areas, the challenges of distorted long-range dependencies from cluttered background distributions and reduced reference information in image domain inevitably rise, which further cause severe performance degradation. To address the challenges, we propose a novel large-portion image inpainting approach, namely the Structure-Guided Synergism Transformer (SyFormer), to rectify the discrepancies in feature representation and enrich the structural cues from limited reference. Specifically, we devise a dual-routing filtering module that employs a progressive filtering strategy to eliminate invalid noise interference and establish global-level texture correlations. Simultaneously, the structurally compact perception module maps an affinity matrix within the introduced structural priors from a structure-aware generator, assisting in matching and filling the corresponding patches of large-proportionally damaged images. Moreover, we carefully assemble the aforementioned modules to achieve feature complementarity. Finally, a feature decoding alignment scheme is introduced in the decoding process, which meticulously achieves texture amalgamation across hierarchical features. Extensive experiments are conducted on two publicly available datasets, i.e., CelebA-HQ and Places2, to qualitatively and quantitatively demonstrate the superiority of our model over state-of-the-arts.

Design of a Multi-Position Alignment Scheme.

The current new type of inertial navigation system, including rotating inertial navigation systems and three-autonomy inertial navigation systems, has been increasingly widely applied. Benefited by the rotating mechanisms of these inertial navigation systems, alignment accuracy can be significantly enhanced by implementing IMU (Inertial Measurement Unit) rotation during the alignment process. The principle of suppressing initial alignment errors using rotational modulation technology was investigated, and the impact of various component error terms on alignment accuracy of IMU during rotation was analyzed. A corresponding error suppression scheme was designed to overcome the shortcoming of the significant scale factor error of fiber optic gyroscopes, and the research content of this paper is validated through corresponding simulations and experiments. The results indicate that the designed alignment scheme can effectively suppress the gyro scale factor error introduced by angular motion and improve alignment accuracy.

Simulation alignment of optical system for space gravitational wave telescope

The performance of telescopes, important components of space interferometry systems, directly affects the accuracy of gravitational wave signals. Space gravitational wave telescopes typically employ an off-axis four-mirror system. When aligned, this system not only has multiple misalignments, but also exhibits interrelated aberrations from various misalignments. These characteristics may lead to difficult alignment of the telescope system as well as significant deviation between the position of the telescope after alignment and the ideal position. To address these issues, first, a sensitivity matrix equation was established between the misalignment of optical components and the Fringe Zernike coefficients. Based on the sensitivity matrix equation, a damping least-squares evaluation function was constructed to reduce the significant deviation between the aligned and ideal positions. Second, a typical optical system of a space gravitational wave telescope was designed, and the sensitivity matrix was calculated. The relationship between the wavefront distortions caused by misalignments in each optical component was examined. To simplify telescope installation, a strategy using secondary mirrors as compensatory elements was proposed. Finally, to verify the effectiveness of the scheme, 200 sets of tolerance files were randomly generated. Based on the evaluation function of the damping least-squares method, a reasonable damping factor was set to limit the solution range of the misalignment, which enabled calculating the secondary mirror compensation amount. Experimental results indicate that after aligning the 200 random telescope files, the root-mean-square wavefront error was reduced to less than 0.0030λ, and the maximum error between the magnification after alignment and the ideal position magnification was only 0.57%, which confirms the feasibility of this alignment scheme.

Progressive Multiple Alignment of Graphs

The comparison of multiple (labeled) graphs with unrelated vertex sets is an important task in diverse areas of applications. Conceptually, it is often closely related to multiple sequence alignments since one aims to determine a correspondence, or more precisely, a multipartite matching between the vertex sets. There, the goal is to match vertices that are similar in terms of labels and local neighborhoods. Alignments of sequences and ordered forests, however, have a second aspect that does not seem to be considered for graph comparison, namely the idea that an alignment is a superobject from which the constituent input objects can be recovered faithfully as well-defined projections. Progressive alignment algorithms are based on the idea of computing multiple alignments as a pairwise alignment of the alignments of two disjoint subsets of the input objects. Our formal framework guarantees that alignments have compositional properties that make alignments of alignments well-defined. The various similarity-based graph matching constructions do not share this property and solve substantially different optimization problems. We demonstrate that optimal multiple graph alignments can be approximated well by means of progressive alignment schemes. The solution of the pairwise alignment problem is reduced formally to computing maximal common induced subgraphs. Similar to the ambiguities arising from consecutive indels, pairwise alignments of graph alignments require the consideration of ambiguous edges that may appear between alignment columns with complementary gap patterns. We report a simple reference implementation in Python/NetworkX intended to serve as starting point for further developments. The computational feasibility of our approach is demonstrated on test sets of small graphs that mimimc in particular applications to molecular graphs.

CEPC survey and alignment

For CEPC such big size machine survey and alignment, several key issues need to be considered: the measurement and data processing method, error accumulation control, the global and relative accuracy, the measurement efficiency, component alignment method and the deformation problem. To these issues an overall survey and alignment scheme was made. The global datums will be established for the measurement and data reduction. A multi-level control network will be built for the global and relative position control, the establish method and measurement method are introduced and the accuracies are estimated. To the component alignment, the schemes of fiducialization, pre-alignment and tunnel alignment are introduced, the deformation problem was analysed and monitor equipment will be applied. To improve the measurement efficiency a new kind of instrument was designed, the prototype has been fabricated and some experiments has been carried out.

Unsupervised Domain Adaptation for Medical Image Segmentation Using Transformer With Meta Attention.

Image segmentation is essential to medical image analysis as it provides the labeled regions of interest for the subsequent diagnosis and treatment. However, fully-supervised segmentation methods require high-quality annotations produced by experts, which is laborious and expensive. In addition, when performing segmentation on another unlabeled image modality, the segmentation performance will be adversely affected due to the domain shift. Unsupervised domain adaptation (UDA) is an effective way to tackle these problems, but the performance of the existing methods is still desired to improve. Also, despite the effectiveness of recent Transformer-based methods in medical image segmentation, the adaptability of Transformers is rarely investigated. In this paper, we present a novel UDA framework using a Transformer for building a cross-modality segmentation method with the advantages of learning long-range dependencies and transferring attentive information. To fully utilize the attention learned by the Transformer in UDA, we propose Meta Attention (MA) and use it to perform a fully attention-based alignment scheme, which can learn the hierarchical consistencies of attention and transfer more discriminative information between two modalities. We have conducted extensive experiments on cross-modality segmentation using three datasets, including a whole heart segmentation dataset (MMWHS), an abdominal organ segmentation dataset, and a brain tumor segmentation dataset. The promising results show that our method can significantly improve performance compared with the state-of-the-art UDA methods.

Band alignment and interfacial charge transfer in sol-gel derived anatase/rutile heterophase TiO2: explaining the synergistic photocatalytic activity.

In this study, we have synthesized nanostructured titanium dioxide (TiO2) photocatalysts under different configurations, viz., anatase, rutile and anatase/rutile heterophase (Eg = 3.14-2.96 eV) through a sol-gel route. The photocatalytic performance of the heterophase samples was better than that of their phase-pure counterparts. Photocatalytic performance was maximum for sample T12H (50% rutile) with a rate constant, k = (7.38 ± 1.48) × 10-2 min-1. This is ca. 47% greater than that estimated via an extrapolation method using the values for phase pure samples indicating synergistic effects. High-resolution transmission electron microscope (HRTEM) analysis and the phase-dependent broadening of core-level X-ray photoelectron spectroscopy (XPS) analysis indicate the formation of well-defined anatase/rutile interfaces. Interfacial charge transfer is the critical factor for this synergistic effect. To analyze the charge transfer pathways, the possible band alignment scheme was analyzed through the precise determination of band alignments from XPS, employing the Ti 2p3/2 core level as a reference. We found a staggered rutile band alignment with the valence band edge of rutile lying above that of anatase, indicating the transfer of photogenerated free electrons from the conduction band of rutile to that of anatase and hole transfer from the valence band of anatase to rutile. These inferences are validated through Mott-Schottky analysis as well.

AAformer: Auto-Aligned Transformer for Person Re-Identification.

In person re-identification (re-ID), extracting part-level features from person images has been verified to be crucial to offer fine-grained information. Most of the existing CNN-based methods only locate the human parts coarsely, or rely on pretrained human parsing models and fail in locating the identifiable nonhuman parts (e.g., knapsack). In this article, we introduce an alignment scheme in transformer architecture for the first time and propose the auto-aligned transformer (AAformer) to automatically locate both the human parts and nonhuman ones at patch level. We introduce the "Part tokens (PARTs)", which are learnable vectors, to extract part features in the transformer. A PART only interacts with a local subset of patches in self-attention and learns to be the part representation. To adaptively group the image patches into different subsets, we design the auto-alignment. Auto-alignment employs a fast variant of optimal transport (OT) algorithm to online cluster the patch embeddings into several groups with the PARTs as their prototypes. AAformer integrates the part alignment into the self-attention and the output PARTs can be directly used as part features for retrieval. Extensive experiments validate the effectiveness of PARTs and the superiority of AAformer over various state-of-the-art methods.

An Interference Alignment Design with Higher Multiplexing Gain by the Reduced Channel Extension

For a K-user Single Input Single Output fading interference channel, Cadambe and Jafar proposed a scheme that achieves K=2 multiplexing gain asymptotically employing interference alignment technique. This scheme is improved by Choi et al. in which for reduced channel extension, higher multiplexing gain is achievable. But the existing designs require an infinite number of channel extension to achieve the upperbound. Our proposed design attains higher multiplexing gain that is arbitrarily close to the upper-bound by using a lot lower number of channel extension in comparison with the other existing interference alignment schemes. On the other hand, due to the great symbol extension reduction of the channel, the computational complexity to achieve K=2 degrees of freedom (desired upper-bound) is also reduced significantly. Moreover, it can be shown that the proposed design achieves higher interference alignment quality. For a limited channel extension, the proposed scheme leads to a better allocation of the signal space at each receiver to the desired and interfered signal. This results to a higher quality of alignment.

Adaptive Beam Alignment and Optimization for IRS-Aided High-Speed UAV Communications

In the unmanned aerial vehicle (UAV) enabled wireless communication system, the rapid channel change caused by high-speed UAV may cause the beam misalignment problem. It is a challenge to fast establish and maintain the communications link between UAV and ground base stations. To address this challenge, an IRS-aided high-speed UAV adaptive beam alignment (IRS-UAV-ABA) scheme utilizing double-layer long short-term memory (DL-LSTM) spatial angle prediction model is proposed to improve the effectiveness and reliability of the UAV communication. Firstly, DL-LSTM enabled spatial angle prediction is modeled to facilitate the calculation of the beam coverage adapted to the UAV real-time flight trajectory. Secondly, the non-convex optimization problem for maximizing the received signal-to-noise ratio of moving UAV is formulated to obtain the suboptimal solutions of the BS’ beamforming vector and the IRS’ phase shift matrix. Finally, with the aforementioned beamforming vectors, the phase shift matrix and the adaptive beam coverage, the adaptive dual-beams alignment from BS to UAV and from IRS to UAV can be achieved by the proposed DL-LSTM-enabled IRS-UAV-ABA algorithm. Simulation results verify that the proposed scheme not only can effectively avoid the beam misalignment caused by the mobility and perturbation of UAV, but also has superior spectrum efficiency and stable beamforming gain.

Interference Alignment in Channel Estimation Using Hybrid Deep Neural Network with Multi-user Propagation Model for 5G in MM Wave Massive MIMO

Massive Multiple Input Multiple Output (MIMO) technologies, as well as higher frequencies utilization, such as MilliMeter Wave (MM Wave) frequencies, should be utilized in the 5G wireless network. This massive transmission will require a higher frequency band and in cellular massive MIMO, frequencies re-use leads to interference in channel estimation. Interference Alignment (IA) is a transmission strategy of the channel with higher capacities of signal-to-noise ratio. This interference needs to be aligned before the transmission of signal, so this research proposes a novel technique in IA by hybridization of deep learning techniques for Heterogeneous Networks (HetNets). The aim of this work is to develop a new interference alignment scheme based on multi-user with a neural network which can be efficient in mitigation of interference of the networks. The model that is proposed properly exploits the interference alignment concept by developing a multi-user group of small cells assisted by various BSs in the heterogeneous networks for MM Wave. The paper focuses on channel estimation by IA for a 5G network that transmits MMWave using neural networks. The neural network is utilized for learning the mapping functions between the received mmWave channels with negligible overhead. Here Hybrid Deep Neural Network is used with Multi-User Propagation (HDNN-MUP) model for IA and demonstrates a practical upper bound for MIMO array sizes that balance throughput and estimation overhead. The experimental analysis is done by utilizing the parameters namely Signal-to-noise ratio (SNR), Bit Error Rate (BER), computational time, energy efficiency, and spectral efficiency.

Construction Control and Monitoring Platform of a Large-Segment Steel Box Girder with Hoisting Installation

The large-segment hoisting construction technology for bridges is increasingly widely used due to its flexibility and efficiency, although it also poses challenges to construction monitoring. Traditional monitoring technology is unitary with low data processing efficiency, making it difficult to meet the accuracy requirements of large-segment hoisting. The application of digital technology has brought about an opportunity for innovation in bridge construction monitoring technology. To address existing challenges and explore digital applications, this paper takes the integral hoisting construction control of the large-segment steel box girder in a large cross-sea bridge as an example, developing an alignment, stress, and temperature monitoring scheme by taking the key points of hoisting construction control into consideration. A monitoring platform was developed, and the workflow of large-segment hoisting construction monitoring is systematically summarized from the viewpoint of practical engineering, which provides a valuable reference for achieving precise and efficient construction monitoring and control in similar projects.

Effects of surface orientations and polarities on the InN/ZnO heterojunctions band offsets: A first-principles study

Band offsets are calculated between InN and ZnO for both polar-c and nonpolar-a interface orientations through hybrid functionals. The pseudomorphic interfaces are considered by assuming whether unstrained InN or ZnO is taken as the substrate. Periodic superlattices satisfying the electron counting rule are used to model the interfaces. By combining separate interface and bulk calculations, the band offsets are obtained through an electrostatic potential alignment scheme. These band offsets are important to the device designs based on the InN/ZnO heterojunctions. Particularly, for the polar c-plane interface heterojunction, the effects of crystal polarities are discussed in detail. The modeling procedures can be directly applied to the theoretical studies of the heterojunctions between nitrides and ZnO.

Multi-Axial Elastic Averaging for Sub-Micron Passive Alignment of Photonic Components

In recent years, heterogeneous integration (HI) has become a game-changing technology for the construction of complex photonic integrated circuits. Comparing to monolithic integration (MI) technique, chips are stacked or add-on during HI. These components are attached to the substrate using an active or passive alignment scheme. As the passive alignment system does not actively search for the optical axis, it allows more die to be bonded per unit time as compared to active alignment. The main challenge is then to obtain consistent sub-micron alignment accuracy during chip manufacturing process to ensure good device performance. To achieve sub-micron alignment accuracy, slide-stop structures with multi-axial elastic averaging coupling designs are implemented. In this work, optical dies are bonded to silicon interposers with two geometrically different slide-stop designs: triangle and rectangle in designs, respectively. Due to an additional surface contact and smaller internal angle for the former, the triangle slide-stop design improves the translational-axes and rotational-axis post-bond alignment accuracy, from (Reference) −0.27 μm ± 2.54 μm @ 3sigma to (slide-stop) −0.09 μm ± 1.37 μm @ 3sigma, (Reference) 1.00 μm ± 3.19 μm @ 3sigma to (slide-stop) 0.39 μm ± 1.29 μm @ 3sigma, and (Reference) 0.06° ± 0.34° @ 3sigma to (slide-stop) 0.0002° ± 0.15° @ 3sigma respectively. Finally, narrower optical power output distribution of P-down bonded lasers with triangle slide-stop over lasers with rectangle slide-stop is demonstrated.