Yafei Hu, Fang Shi, Zheng Liu et al. · 2018
Underground pipelines are subject to severe distress from the surrounding expansive soil. To investigate the structural response of water mains to varying soil movements, field data, including pipe wall strains in situ soil water content, soil pressure and temperature, was collected. The research on monitoring data analysis has been reported, but the relationship between soil properties and pipe deformation has not been well-interpreted. To characterize the relationship between soil property and pipe deformation, this paper presents a super learning based approach combining feature selection algorithms to predict the water mains structural behavior in different soil environments. Furthermore, automatic variable selection method, e.i. recursive feature elimination algorithm, were used to identify the critical predictors contributing to the pipe deformations. To investigate the adaptability of super learning to different predictive models, this research employed super learning based methods to three different datasets. The predictive performance was evaluated by R-squared, root-mean-square error and mean absolute error. Based on the prediction performance evaluation, the superiority of super learning was validated and demonstrated by predicting three types of pipe deformations accurately. In addition, a comprehensive understand of the water mains working environments becomes possible.
Cited by 1
Wei‐Sheng Lai, Jia‐Bin Huang, Narendra Ahuja et al. · IEEE Transactions on Pattern Analysis and Machine Intelligence · 2018
Convolutional neural networks have recently demonstrated high-quality reconstruction for single image super-resolution. However, existing methods often require a large number of network parameters and entail heavy computational loads at runtime for generating high-accuracy super-resolution results. In this paper, we propose the deep Laplacian Pyramid Super-Resolution Network for fast and accurate image super-resolution. The proposed network progressively reconstructs the sub-band residuals of high-resolution images at multiple pyramid levels. In contrast to existing methods that involve the bicubic interpolation for pre-processing (which results in large feature maps), the proposed method directly extracts features from the low-resolution input space and thereby entails low computational loads. We train the proposed network with deep supervision using the robust Charbonnier loss functions and achieve high-quality image reconstruction. Furthermore, we utilize the recursive layers to share parameters across as well as within pyramid levels, and thus drastically reduce the number of parameters. Extensive quantitative and qualitative evaluations on benchmark datasets show that the proposed algorithm performs favorably against the state-of-the-art methods in terms of run-time and image quality.
Cited by 920
Wen-Xuan Long, Rui Chen, Marco Moretti · IEEE Transactions on Vehicular Technology · 2022
Multi-user orbital angular momentum (MU-OAM) wireless backhaul systems require precise alignment of transmitted and received OAM beams in line-of-sight (LoS) scenarios. Therefore, a crucial feature for a macro base station (MBS) is to accurately obtain the angle of arrival (AoA) of OAM beams from the small base stations (SBSs) in the system. Considering the limitations of existing OAM-based AoA estimations, we propose a novel super-resolution AoA estimation method, denoted as the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">recursive estimation of signal parameters via rotational invariance techniques (ESPRIT)</i> method. The proposed technique is fundamentally different from traditional ESPRIT algorithm: It is not based on the displacement-invariant array structure, but on the recursive characteristic of Bessel functions in OAM signals. With this method, the inter-mode interference of the LoS MU-OAM system can be significantly suppressed by less training overhead, and the performance is close to that of the ideal system.
Cited by 12Open access
Shannon D. Blunt, Tszping Chan, K. Gerlach · 2008
A new approach for spatial direction-of-arrival (DOA) estimation is developed based on the minimum mean-square error (MMSE) framework. Unlike many traditional DOA estimators, the MMSE approach, denoted as Re-Iterative Super-Resolution (RISR), does not employ spatial sample covariance information which may significantly degrade DOA estimation if spatially-separated sources are temporally correlated. Instead, RISR is a recursive algorithm that relies on a structured signal covariance matrix comprised of the set of possible spatial steering vectors each weighted by an associated power estimate from the previous iteration. Furthermore, RISR can naturally accommodate prior information on spatially colored noise, does not require knowledge of the number of sources, and can also exploit multiple time samples in a non-coherent manner to improve performance. For low to moderate time sample support, RISR is demonstrated to provide super-resolution performance superior to MUSIC and spatially-smoothed MUSIC.
Cited by 30
Jiawei Luo, Yongchao Zhang, Yongwei Zhang et al. · 2021
Airborne radar high-squint looking imaging is an important research for remote sensing. The traditional Doppler beam sharpening based on fast Fourier transform (FFT) has good real-time performance but low cross-range resolution. Many super-resolution methods have been proposed to enhance the cross-range resolution for airborne radar. However, these methods generally adopt the batch processing mode with high computational complexity and high memory usage, which lead to poor real-time performance. This paper proposes an online super-resolution imaging approach for airborne scanning radar based on sliding window recursive least square (SWRLS) algorithm. The current scattering estimation can be derived recursively through downdating and updating. The proposed method effectively improves the cross-range resolution as well as the real-time performance and memory occupancy, which is beneficial to high-quint continuous realtime imaging for airborne radar. Simulation results are given to demonstrate the effectiveness of the proposed method.
Cited by 2
Rui Chen, Wenchao Li, Jianyu Yang et al. · IEEE Transactions on Aerospace and Electronic Systems · 2025
By combining the echoes of two channels in azimuth, and utilizing the super-resolution algorithm for ill-posed problems, the left/right ambiguity in forward-looking synthetic aperture radar (SAR) can be resolved. However, due to the large size of the echo matrix and the need for multiplication and inversion operations, a heavy computational burden is usually encountered. In this paper, based on the time-frequency-space steering matrix, an accelerated left/right ambiguity resolving and angular super-resolution scheme for dual-channel forward-looking SAR imaging is proposed. In the scheme, the Hermitian Toeplitz Block Toeplitz (HTBT) property is exploited to accelerate the construction of the sample covariance matrix. Then, based on the Gohberg Semencul (GS) decomposition theorem and the 2D Levinson Durbin recursive algorithm, the inversion process of the high dimensional covariance matrix is transformed into a large number of low dimensional matrix inversion operations. Finally, based on the Toeplitz multiplication and the trigonometric polynomial expansion theories, the multiplication operations involved between the inverse sample covariance matrix and matrices or vectors are realized by FFT. Simulation and measured data demonstrate that this scheme is more efficient than existing algorithms while maintaining imaging performance.
Cited by 1
Sébastien Lablanche, gerard Lablanche · Journal of Current Trends in Computer Science Research · 2023
In this paper, we explain a process of super-resolution reconstruction allowing to increase the resolution of an image. The need for high-resolution digital images exists in diverse domains, for example, the medical and spatial domains. The obtaining of high-resolution digital images can be made at the time of the shooting, but it is often synonymic with important costs because of the necessary material to avoid such costs, it is known how to use methods of super-resolution reconstruction, consisting of one or several low-resolution images to obtain a high-resolution image. The American patent US 9208537 describes such an algorithm. A zone of one low-resolution image is isolated and categorized according to the information contained in pixels forming the borders of the zone. The category of it zone determines the type of interpolation used to add pixels in the aforementioned zone, to increase the neatness of the images. It is also known how to reconstruct a lowresolution image and a high-resolution image by using a model of super-resolution reconstruction whose learning is based on networks of neurons and on an image or a picture library. The demand of Chinese patent CN 107563965 and the scientist publication “Pixel Recursive Super Resolution “, R. Dahl, M. Norouzi, and J. Shlens propose such methods. The aim of this paper is to demonstrate that it is possible to reconstruct coherent human faces from very degraded pixelated images with a very fast algorithm, faster than the compressed sensing (CS) algorithm, easier to compute and without deep learning, so without important information technology resources, i.e. a large database of thousand of training images ( https://arxiv. org/pdf/2003.13063.pdf). This technological breakthrough has been patented in 2018 with the demand of French patent FR 1855485 (https://patents.google.com/patent/FR3082980A1, see the HAL reference https://hal.archives-ouvertes.fr/hal01875898v1)
Cited by 0Open access
Alvaro Lopez Paredes, Miguel Heredia Conde, Thoriq Ibrahim et al. · Zenodo (CERN European Organization for Nuclear Research) · 2023
Time-of-Flight cameras, especially the ones featuring sensors built upon macro-pixel architectures, are generally characterized by lower resolutions than other imaging modalities. In this work, we propose a practical single-frame super-resolution scheme to surpass this limitation by exploiting the architecture of a laboratory-designed macro-pixel ultra-high-speed image sensor. In addition, we propose a two-step sparsity-aware greedy algorithm for the recovery of the signal in the time (depth) domain. The main contribution of our algorithm is the introduction of a preliminary screening step to define a set of feasible support candidates in which we replace the demodulation functions, the assembly of which yields the sensing matrix, by the theoretical binary codes they are built upon. We empirically demonstrate that our algorithm improves the recovery performance with respect to Order Recursive Matching Pursuit for sparsity <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\leq 2$</tex> .
Cited by 4Open access
Gavin E. Mendel-gleason, Geoff Hamilton · 2010
Abstract. It has been long recognised that partial evaluation is related to proof normalisation. Normalisation by evaluation, which has been presented for theories with simple types, has made this correspondence formal. Recently Andreas Abel formalised an algorithm for normalisation by evaluation for System F. This is an important step towards the use of such techniques on practical functional programming languages such as Haskell which can reasonably be embedded in relatives of System Fω. Supercompilation is a program transformation technique which performs a super-set of the simplifications performed by partial evaluation. The focus of this paper is to formalise the relationship between supercompilation and normalisation by evaluation for System F with recursive types and terms. 1
Cited by 0
Low, Dunstan · Zenodo (CERN European Organization for Nuclear Research) · 2025
This section formalizes a theoretical extension to the Single Point Super Projection – Single Sphere Cosmology (SPSP–SSC), emphasizing scale invariance, where quantum and cosmological phenomena emerge as non-hierarchical projections from a single sphere (). Matter arises from horizon-external interactions forming electromagnetic (EM) shells, with particles encoding recursive cosmologies. A ratio cone, driven by SPS qubit spin and elliptic sorting, unifies mass, EM, and gravity. Localized phenomena (e.g., planetary orbits) manifest in a holographic, multidirectional 3D spacetime within an infinitely nested node field of autonomous yet connected projections. Gravitational vectors act as colliding branches, with the elliptic constraint as a sorting algorithm. A QuTiP simulation visualizes nodes, vectors, and collisions, enhancing rigor and clarity.
Cited by 0Open access
Burgin, Mark, Klinger, Allen · eScholarship (California Digital Library) · 2004
This paper extends traditional models of machine learning beyond their one-level structure by introducing previously obtained problem knowledge into the algorithm or automaton involved. Some authors studied more advanced than traditional models that utilize some kind of predetermined knowledge, having a two-level structure. However, even in this case, the model has not reflected the source and inherited properties of predetermined knowledge. In society, knowledge is often transmitted from previous generations. The aim of this paper is to construct and study algorithmic models of learning processes that utilize predetermined or prior knowledge. The models use recursive, subrecursive, and super-recursive algorithms. Predetermined knowledge includes: a text description, activity rules (e.g., for cognition), and specific structured personal or social memory. Algorithmic models represent these three forms as separate structured processing systems: automata with (1) advice; (2) structured program; and (3) structured memory. That yields three basic models for learning systems: polynomially bounded turing machines, Turing machines, and inductive Turing machines of the first order. (C) 2003 Elsevier B.V. All rights reserved.
Cited by 0Open access
Guoteng Zhang, Peng Ran, Zehua Zhang · Scientific Reports · 2025
This paper addresses the high-frequency oscillations issue commonly associated with the sliding mode observer (SMO) algorithm in sensorless control of surface-mounted permanent magnet synchronous motor (SPMSM) and the problem of parameter variation caused by magnetic saturation and temperature rise during motor operation. A novel information acquisition optimized recursive least squares (IAORLS) algorithm, which incorporates an IAO optimized forgetting factor, is proposed to identify motor parameters. The identified parameters are utilized in conjunction with the Super Twisting Algorithm Sliding Mode Observer (STA-SMO) for motor control. By discretizing the stator voltage equations of the SPMSM via the forward Euler method, IAORLS enables efficient parameter identification. These parameters are then integrated into the STA-SMO, allowing the observer to receive real-time updates, thereby preventing performance degradation due to parameter mismatches. Additionally, the method reduces high-frequency oscillations when the motor's extended back electromotive force (back-EMF) is observed, improving the accuracy of rotor position and speed estimation. Simulations and experimental results confirm the effectiveness and reliability of the proposed approach.
Cited by 3Open access
Low, Dunstan · Zenodo (CERN European Organization for Nuclear Research) · 2025
This section formalizes a theoretical extension to the Single Point Super Projection – Single Sphere Cosmology (SPSP–SSC), emphasizing scale invariance, where quantum and cosmological phenomena emerge as non-hierarchical projections from a single sphere (). Matter arises from horizon-external interactions forming electromagnetic (EM) shells, with particles encoding recursive cosmologies. A ratio cone, emerging from SPS qubit spin and elliptic sorting, unifies mass, EM, and gravity. Localized phenomena (e.g., planetary orbits) manifest in a holographic, multidirectional 3D spacetime within an infinitely nested node field of autonomous yet connected projections. Gravitational vectors act as colliding branches, with the elliptic constraint functioning as a sorting algorithm. This preserves SPSP–SSC’s lock to General Relativity (GR), the Standard Model (SM), and Quantum Mechanics (QM) while offering falsifiable predictions.
Cited by 0Open access
Peter F. Stadler · Interface Focus · 2021
Alignments of discrete objects can be constructed in a very general setting as super-objects from which the constituent objects are recovered by means of projections. Here, we focus on contact maps, i.e. undirected graphs with an ordered set of vertices. These serve as natural discretizations of RNA and protein structures. In the general case, the alignment problem for vertex-ordered graphs is NP-complete. In the special case of RNA secondary structures, i.e. crossing-free matchings, however, the alignments have a recursive structure. The alignment problem then can be solved by a variant of the Sankoff algorithm in polynomial time. Moreover, the tree or forest alignments of RNA secondary structure can be understood as the alignments of ordered edge sets.
Cited by 8Open access
P. N. Bashly, V. G. Denisenko, S. E. Mishchenko et al. · Antennas · 2025
Two algorithms for singular value decomposition of the correlation matrix have been proposed for the implementation of angular super-resolution and the construction of the spatial spectrum of a digital antenna array. The first algorithm is based on the nonlinear iterative partial least squares (NIPALS) method, which is usually used to calculate the principal components of large amounts of data. In contrast to the main NIPALS algorithm, the possibility of recursively specifying initial approximations of eigenvectors based on the results of signal processing by a digital antenna array has been used. To process each subsequent correlation matrix, the eigenvectors of the previous processing phase have been used. The second algorithm is based on the parallel Jacobi algorithm and is distinguished by the use of analytical relations to calculate complex rotation matrices. It has been shown that the recursive assignment of the initial approximations of the eigenvectors in the first algorithm made it possible to significantly reduce the required number of iterations. This makes the first algorithm potentially more efficient than the Jacobi algorithm. In turn, the Jacobi algorithm is more resistant to limiting the length of the mantissa. A simulation has been performed confirming the operability and convergence of the proposed MUSIC algorithms based on two different singular expansions with the resolution of two closely spaced sources.
Cited by 0
Kenneth B. Pasiah, Lih‐Yuan Deng, Dale Bowman et al. · 2021
Pseudo-random numbers (PRNs) are the basis for almost any statistical simulation and this depends largely on the quality of the pseudo-random number generator (PRNG) used. In this study, we used some results from number theory to propose an efficient method to accelerate the computer search of super-order maximum-period multiple recursive generators (MRGs). We conduct efficient computer searches and identify many efficient and portable MRGs of super-orders, 40751, 50551, and 50873; which respectively have equi-distribution property up to 40751, 50551, and 50873 dimensions, and period lengths of approximately 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">380278.1</sup> , 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">471730.6</sup> , and 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">474729.3</sup> . Using the generalized Mersenne prime algorithm, we extend some known results of some efficient, portable and maximum-period MRGs. In particular, the DX/DL/DS/DT large-order generators are extended to super-order generators. An extensive empirical evaluation shows that these generators behave well when tested with the stringent Small Crush and Crush batteries of the TestU01 package.
Cited by 0
Zhe Dong, Kai Liu, Shipeng Wang · Drones · 2022
Unmanned aerial vehicles (UAVs) have been widely applied over the past decades, especially in the military field. Due to the unpredictability of the flight environment and failures, higher requirements are placed on the design of the control system of the fixed-wing UAV. In this study, a sliding mode disturbance observer-based (SMDO) adaptive dynamic inversion fault-tolerant controller was designed, which includes an outer-loop sliding mode observer-based disturbance suppression dynamic inversion controller and an inner-loop real-time aerodynamic identification-based adaptive fault-tolerant dynamic inversion controller. The sliding mode disturbance observer in the outer-loop controller was designed based on the second-order super-twisting algorithm to alleviate chattering. The aerodynamic identification in the inner-loop controller adopts the recursive least squares algorithm to update the aerodynamic model of the UAV online, thereby realizing the fault-tolerant control for the control surface damage. The effectiveness of the proposed SMDO enhanced adaptive fault-tolerant control method was validated by mathematical simulation.
Cited by 17Open access
Ruitao Li, Jiawei Luo, Yin Zhang et al. · Remote Sensing · 2025
The sparse iterative covariance-based estimation (SPICE) method has recently gained significant attraction in the field of scanning radar super-resolution imaging because of its angular resolution enhancement capability. However, it is unable to preserve the target profile, and the estimator is constrained by high computational complexity and memory consumption. In this paper, a grid-updating split SPICE-TV algorithm is presented. The method allows for the efficient updating of reconstruction results with both contour and resolution, and a recursive grid-updating implementation framework of the split SPICE-TV has the capability to reduce the computational complexity. First, the scanning radar angular super-resolution problem is transformed into a constrained optimization problem by simultaneously employing sparse covariance fitting criteria and TV regularization constraints. Then, the split Bregman method is employed to derive an efficient closed-form solution to the problem. Ultimately, the matrix inversion problem is transformed into an online iterative equation to reduce the computational complexity and memory consumption. The superiority of the proposed method is verified by simulation and experimental data.
Cited by 2Open access
Marios C. Papaefthymiou, Jiyoun Kim · Deep Blue (University of Michigan) · 2006
Partitioning can speed up overlong VLSI design processes by enabling process parallelization. To achieve high parallel processing efficiency, partitioning schemes should minimize imbalance or interactions among partitions by relying on the specific characteristics of the given processes. In this dissertation, we present two partitioning methodologies for parallelizing post-placement VLSI optimization processes. Unlike pre-placement VLSI design tasks, post-placement processes manipulate circuit elements that have been assigned to physical spaces on the chip die. Consequently, physically neighboring circuit components may interact even when they are independent in the circuit netlist. To achieve high parallelism, partitioning methods must therefore take into consideration both logic and geometric interactions of circuit components. First we present a linear-time bipartitioning algorithm for the <bold> 2-weight geometric load balancing</bold> problem. This algorithm uses an L-shaped separator to bipartition a 2-weighted graph in 2-dimensional space so that each partition is spatially connected and each of the two weights is distributed with asymptotically vanishing relative imbalance. We also give an efficient heuristic for 2<italic><super>p</super></italic>-way partitioning based on the recursive application of our bipartitioning algorithm. Our heuristic runs in <italic>O</italic>(<italic>pn</italic> + n ·2<italic><super>p</super></italic>) time, where <italic> n</italic> is the graph size. In experiments with geometric graphs obtained from placed circuits, our partitioning heuristic achieves good balancing, fast runtime, and netcut sizes comparable to other partitioners. The second contribution of this dissertation is a <bold>multi-session partitioning</bold> scheme. This partitioning technique divides problem tasks into multiple sessions of parallel processes, so that interprocessor communication is entirely removed during each session. The algorithm is especially useful for parallelizing processes whose tasks are heavily connected to each other and may result in high communication overhead in parallel processing. We have developed a C++ version of our multi-session partitioning scheme and applied it to two post-placement VLSI processes with heavily overlapping tasks: electrical violation correction and post-placement timing optimization. In the parallelization of electrical violation correction, our scheme achieves speedups of up to 2.3x, dynamically utilizing 1--6 processors. For the process of post-placement timing optimization, we have developed a path-based buffer insertion algorithm and software. Our partitioning scheme speeds up the software execution time by up to 4.9x over serial processing, while dynamically utilizing 1--8 processors.
Cited by 1
Jiaqing Liu, Yi Li, Dong Liang et al. · 2023
Due to the limitation of the detector, the spatial resolution of the polarization image obtained by space-modulated full-polarization computed imaging is low. In general, there are a lot of valuable high-frequency components in low resolution space, but it is difficult to reconstruct high frequency information by deep learning method. Through the research of the previous super-resolution reconstruction method based on deep learning, it is found that it is difficult to obtain better lifting effect only by stacking residual blocks to build a deeper network. A polarization imaging bi-attention recursive residual super-resolution reconstruction method is proposed. In the network structure, bi-attention recursive residual group is used as the deep feature extraction module, and the hybrid bi-convolutional attention module contained in this module is used to adaptively learn image features of different channels and different transformation Spaces in the deep network. In order to provide the concentrated learning efficiency of high-frequency information, jump connections are introduced into the network structure, and the shallow feature extraction and reconstruction of images are completed by a convolution layer and a sub-pixel convolution algorithm respectively. The experimental verification is carried out under the actual imaging system and simulation data and compared with other methods in terms of visual effect and quantitative results. Experimental results verify that the proposed method can effectively reconstruct the contour structure and texture details of the detection target in subjective image quality and is superior to the comparison method in objective evaluation index.
Cited by 0