Irregular Sampling Research Articles

Novel Trajectory Representation Learning Method and Its Application to Trajectory-User Linking

With the widely used mobile phones, the user's trajectory data can be easily collected by the base station. The trajectory representation learning is the upstream task of trajectory modeling in the transportation systems. And excellent pretraining features will greatly improve the downstream model. However, the trajectory data collected by the base station has a lot of missing values, and the trajectory data contains the coordinate internal features and coordinate context features resulting in the traditional embedding methods in the real space being not suitable for the trajectory data. Therefore, this article proposes novel trajectory representation learning method in the complex space based on the Skipgram (TRMC+Skipgram). The trajectory context is represented by a graph to solve the irregular sampling, where the nodes of the graph are coordinates and the weights of the edges are proportional to the number of co-occurrences of the two nodes. Meanwhile, the coordinates are embedded into the complex space for representing both the coordinate context feature and coordinate internal feature. Through the linking prediction evaluated, the TRMC+Skipgram achieves the state-of-the-art results and yields a 3.7% improvement compared with the best baseline on the AUC metric. In order to verify that the robust pretraining vectors can improve the effect of downstream tasks, the trajectory-user linking model (TULM) is constructed based on TRMC+Skipgram, namely TULM+TRMC+Skipgram. In the task of user-trajectory linking, the TULM+TRMC+Skipgram achieves the state-of-the-art result and yields 1.7% improvement compared with best baseline on the macro- F <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> metric.

Dynamic Point Cloud Inpainting via Spatial-Temporal Graph Learning

The maturity of depth sensors and laser scanning techniques has enabled the convenient acquisition of 3D dynamic point clouds—one natural representation of 3D objects/scenes in motion, leading to a wide range of applications such as immersive tele-presence, autonomous driving, augmented and virtual reality. Nevertheless, dynamic point clouds usually exhibit holes of missing data, thus inpainting is crucial to the subsequent rendering or downstream understanding tasks. Dynamic point cloud inpainting has been largely overlooked so far, which is also quite challenging due to the irregular sampling patterns both in the spatial domain and temporal domain. To this end, we propose an efficient dynamic point cloud inpainting method based on a learnable spatial-temporal graph representation, exploiting both the second-order inter-frame coherence and the intra-frame self-similarity. The key is the second-order inter-frame coherence that enforces the consistent flow in 3D motion over time, for which we search the temporal correspondence in consecutive frames for the same underlying surface by the point-to-plane distance and represent the correlation between them via temporal edge weights in the graph. Based on the second-order inter-frame coherence and intra-frame self-similarity, we formulate dynamic point cloud inpainting as a joint optimization problem of the desired point cloud and underlying spatial-temporal graph, which is regularized by consistency in the temporal edge weights and smoothness in the spatial domain. We analyze and reformulate the optimization, leading to an efficient alternating minimization algorithm. Experimental results show that the proposed approach outperforms several competing methods significantly, both on synthetic holes and real holes.

Deep Prior-Based Unsupervised Reconstruction of Irregularly Sampled Seismic Data

Irregularity and coarse spatial sampling of seismic data strongly affect the performances of processing and imaging algorithms. Therefore, interpolation is a usual preprocessing step in most of the processing workflows. In this work, we propose a seismic data interpolation method based on the deep prior paradigm: an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ad hoc</i> convolutional neural network is used as a prior to solve the interpolation inverse problem, avoiding any costly and prone-to-overfitting training stage. In particular, the proposed method leverages a multiresolution U-Net with 3-D convolution kernels exploiting correlations in cubes of seismic data, at different scales in all directions. Numerical examples on different corrupted synthetic and field data sets show the effectiveness and promising features of the proposed approach.

Effect of Ni ions on thermal condensation kinetics and morphology of g-C3N4 nanosheets and their photo- and electro-chemical activity evolution

The effect of Ni ions on the thermal condensation kinetics, morphology, microstructure and photoelectric properties of layered g-C3N4 were tested. Ni ions were first incorporated into melamine before thermal polymerization. The morphologies of g-C3N4 from nanosheets to irregular tubular structure was created because of the effect of Ni atoms. The photo- and electro-chemical performance depended strongly on the morphologies of g-C3N4. Irregular tubular sample NCN675 with multiple scattering effect of light exhibited fast degradation for Rh B under visible light with a maximum degradation rate constant of 0.046 min−1. Electrocatalysis test indicates that NCN680 as N-doped carbon nanotubes with excellent electrocatalytic hydrogen and oxygen evolution rate (HER and OER), in which Tafel slopes of HER and OER are 111 and 107 mV/dec respectively. In addition, charge discharge test of supercapacitor confirms that NCN680 is the most suitable electrode material for supercapacitors and have larger specific capacitance (SC = 203.95F/g) when its current density is 1A/g. This study provides an important supplement for structure and performance transformation of carbon nitride compounds catalyzed by Ni.

A non-linear elastic approach to study the effect of ambient humidity on sandstone

We demonstrate that strong elastic pump wave pulses soften sandstone more in humidified conditions than they do in dry conditions and that this effect is repeatable and reversible. We assess these changes via the non-linear interactions of a strong pump wave with a weaker probe wave. We find that there is an exponential time constant (τ≈13days) associated with this process that is independent of the amplitudes of the pump and the probe, the phase delay between the two waves (the time between transmission of the pump and probe waves), the sampling rate, and whether the sample is being dried or humidified. We demonstrate that the humidity-dependent differences in the amount of softening are induced by only a very small amount of absorbed water vapor and argue that this water is intercalated within clay particles. We also show that our pump–probe experiments detect these humidity-dependent differences in the amount of softening easily and repeatably using an experimental design that does not rely on resonance conditions. This means that, in principle, our experiments could be more easily generalized to other experimental geometries to investigate non-linear elastic properties in complex or irregular sample geometries. Our method and findings have potential relevance in oil and gas exploration, civil engineering, and understanding of the mechanism of earthquakes.

A type of X-ray diffractometer with adaptive X-ray spot sizes

In order to realize micron scale to millimeter scale phase structure analysis, as well as accurate phase structure analysis of surface uneven samples, X-ray diffractometer named Hawk-II, which can adaptively adjust the diameter of irradiated X-ray beam spot according to the diameter of internal tangential circle at the measured point, is developed by combining X-ray diffraction technology, CCD camera imaging technology and slightly-focusing ploycapillary X-ray control technology. The X-ray source system, six-dimensional linkage motion system, CCD camera, detection system and control system based on LabVIEW are the main components of the Hawk-II. Compared with the 3°–5° divergence of the conventional X-ray source, the divergence of the X-ray emitted by the slightly-focusing polycapillary X-ray optics is only about 0.15° and also the intensity within the beam spot range is dozens of times stronger. Therefore, the shift of peak position will not appear due to the pores, curvature or uneven surface of the sample, when Hawk-II is used to analyze the samples with irregular surface. The diffraction pattern of the uneven Ren Min Bi five-cent coin are collected in the Hawk-II and PANalytical X-Pert Pro MPD conventional X-ray diffractometer respectively. By comparing the analysis results, it is found that the diffraction peaks measured by the X-Pert Pro MPD are shifted seriously, with a maximum deviation angle of 0.52°. While the diffraction peaks detected by the Hawk-II are basically consistent with the data from the standard PDF card, which verifies the advantages of the analysis of irregular samples by the Hawk-II. In order to explore the difference between different beam spots used for analysis at the same point, red and green porcelain fired in Qing dynasty and GaAs-based Cu and Fe plated films are analyzed by the Hawk-II. It is found that when the samples are relatively uniform, the intensities of diffraction peaks of different beam spots are relatively close, while when the samples are not uniform, the diffraction peaks vary greatly. Especially, some microcrystalline phases can be detected only with large beam spots. In addition, to verify the adaptive functionality of the Hawk-II, a bronze from the Western Han Dynasty, with different rust spots on it, is tested. It is found that the Hawk-II can adjust the beam spot size according to the different corrosion points, making the irradiation area coincide with the area to be analysed and the phase structure detected more accurately. Therefore, the Hawk-II is a general purpose X-ray diffractometer, which has the analytical capability from micron scale to millimeter scale and the energy dispersive X-ray fluorescence analysis function. Moreover, it has the advantages of the accurate analysis of irregular samples, fast detection speed, simple operation, etc. Based on the above analysis, the Hawk-II will be widely used in different fields.

(Invited) Innovative 3D-Printed Flow-Type Electrochemical Microcell for Ultra-High Spatial Resolution Electrochemistry

Advances in 3D-printing technology have been used to design and develop a customizable flow-type electrochemical microcell that provides analytical resolution comparable to that of conventional glass capillary-based electrochemical microcells. In addition to being highly reproducible compared to the fabrication of glass capillaries, the advantages of this design are increased durability, decreased production costs, and better control over design parameters while maintaining ultra-high spatial resolution. The design is a three-component microcell that combines 3D printing with conventional machining to allow rapid prototyping and testing of flow designs and apertures sizes while maintaining a consistent foundation upon which the design can be built. The key benefit of the three-component design is the ability to replace parts as needed without complete disassembly. The design and application of two classes of flow-type electrochemical microcell, high-resolution (~100 µm spot size) and ultra-high resolution (~10 µm spot size), will be presented. Additionally, the proposed prototype introduces the practicality of a six-axis robot arm for positioning the microcell and is advantageous over a conventional three-axis gantry system allowing for the analysis of irregular sample geometries. The microcell prototype was tested and evaluated to demonstrate repeatability in use, accuracy in spatial resolution and experimental results, and capability in the analysis of various materials and surface finishes. The ease of use and endurance of the design is also demonstrated through repeated use. This microcell has the potential for applications in corrosion, electrodeposition, and electroanalysis, amongst others.Acknowledgements: This work was supported by the United States Nuclear Regulatory Commission (NRC) under contract NRC-HQ-84-16-G-0033. Ms. Nancy Hebron-Israel serves as the program manager.

3-D and 5-D reconstruction ofPreceiver functions via multichannel singular spectrum analysis

SUMMARYThe receiver function (RF) method is fundamental in assessing mantle seismic discontinuity depths and reflectivities. Most of the current approaches rely on phase equalization, though in many applications, high levels of incoherent noise, incomplete and irregular sampling customarily interfere with the analysis of weak secondary phases. In recent years, advancements in the field of multidimensional seismic data processing have triggered a shift in interest towards its application to RFs, specifically to single station gathers that depend on a single spatial dimension. Our work generalizes the application of singular spectrum analysis to RFs that rely on two and four spatial dimensions recorded by dense seismic arrays. We adopt a multidimensional signal processing approach known as multichannel singular spectrum analysis. We develop a strategy to assemble and enhance 3-D and 5-D seismic volumes via matrix rank reduction and a reinsertion algorithm to simultaneously suppress random noise, retrieve absent observations and boost identifiability of secondary conversions. We provide informative synthetic examples to gain insight into the effectiveness and limitations of our approach. In the real data example, we improve weak conversions from the mantle transition zone (MTZ) recorded by the USArray in the Yellowstone area. The reconstruction algorithm accurately recovers the timing and polarity of conversions associated with the 410-, 520- and 660-km seismic discontinuities. Our investigation shows that the simultaneous processing of several spatial variables expedites signal restoration, particularly in directions where large recording gaps exist due to a lack of earthquakes, which aids the mapping and identification of the MTZ interfaces. This study presents a theoretical/practical framework for the reconstruction of multidimensional RF data, and its full potential can be exploited with dense acquisition available from the emerging seismic nodal arrays to improve passive seismic imaging.

Intelligent Detection for Key Performance Indicators in Industrial-Based Cyber-Physical Systems

Intelligent anomaly detection for key performance indicators (KPIs) is important for keeping services reliable in industrial-based cyber-physical systems (CPS). However, it is common in practice for various KPI sampling strategies to be utilized. We experimentally verify that anomaly detection is highly sensitive to irregular sampling, and accordingly go on to investigate low-cost anomaly detection for large-scale irregular KPIs. Irregular KPIs can be classified into four types: equal interval and unequal quantity (EIUQ) KPIs, unequal interval (UI) KPIs, unequal interval with equal duration (UIED) KPIs, and segmented irregular KPIs. In this article, we propose an anomaly detection framework based on these irregular types. Moreover, to handle the various lengths and phase shifts among EIUQ KPIs, we propose a normalized version of unequal cross-correlation, which slides the KPIs to enable finding the most similar position. To avoid high computational costs, we analyze the low-rank feature of KPIs data and propose a matrix factorization-based alignment algorithm for UIED KPIs; this algorithm treats UIED KPIs as an incomplete matrix and recovers the KPIs to align them before performing anomaly detection. Extensive simulations using three public datasets and two real-world datasets demonstrate that our algorithm can achieve a larger F1-score than Minkowski distance and less time than dynamic time warping distance.

Application of Nonuniform FFT to Spherical Near-Field Antenna Measurements

This article investigates on the use of nonuniform fast Fourier transform (NUFFT) to solve spherical near-field transformation problems. Traditional postprocessing algorithms make use of conventional FFT for the transformation of measured grids with uniform sampling. More recently, the processing of irregular sampling grids has been introduced by means of matrix-based algorithms. The benefits of both type of algorithms are combined in this article with the implementation of the spherical wave expansion (SWE) summation and its adjoint with NUFFT. This allows the processing of fields with the efficiency of the FFT approaches and the versatility of the matrix operators. To this end, the NUFFT will be introduced and it will be shown how it can be implemented to accelerate the multidimensional summations required in the SWE calculations. Numerical investigations will be introduced to demonstrate the computational gains achieved with negligible approximation errors. Finally, the NUFFT will be tested on three different spherical near-field postprocessing problems: far-field transformation of near fields measured with spiral scanning, phaseless measurements on spherical surfaces with different sampling grids, and antenna diagnostics.

Activities of Daily Living Recognition With Binary Environment Sensors Using Deep Learning: A Comparative Study

The power of end-to-end deep learning techniques to automatically learn latent high-level features from raw signals has been demonstrated in numerous application fields, however, few studies systematically investigate how to properly encode the time-series firings of binary environment sensors that typically work in an event-triggering scheme and have irregular sampling rates for in-home human activity recognition. To this end, we here propose two different methods to process the streaming sensor readings and accordingly evaluate their combinations with deep learning models. Specifically, we divide the multichannel sensor events into segments and encode each segment into either a vector or two-dimensional matrix. Particularly, three different feature representations are presented for the vector form. Afterwards, we combine the encoded features with four typical deep learning models and optimize corresponding activity recognizers to study their sensitivity to different feature encodings. Furthermore, we include seven commonly used shallow classification models for comparison purposes. Finally, we conduct extensive experiments on three publicly available smart home datasets. Results indicate that the performance of both deep learning and shallow models is closely associated with the raw signal encodings and demonstrate the superiority of one-dimensional convolutional neural networks over its competitors in terms of generalization across scenarios. Besides, we preliminarily explore the influence of NULL class on an activity recognizer and experimentally show its negative impact on overall accuracy, enlightening relevant studies to consider it in developing a practical activity recognition system.

Improved etching uniformity using equivalent electrodes on an unconventional, irregular membrane optical element for large aperture diffractive optical telescopes.

The etching uniformity of diffractive membrane optical elements with an irregular shape was investigated. A deteriorative uniformity of electron number density and electron temperature was found according to finite element analysis of plasma discharge. A designable equivalent electrode was proposed to weaken the influence of introducing the unconventional, irregular sample. Improved uniformity of etching depths was demonstrated experimentally, assisting by the designable equivalent electrode. The demonstration of the designable equivalent electrode provides a beneficial solution for the fabrication of unconventional optical elements and an effective means for adjusting and controlling plasma characteristics.

Anomaly detection and event mining in cold forming manufacturing processes

Predictive maintenance is one of the main goals within the Industry 4.0 trend. Advances in data-driven techniques offer new opportunities in terms of cost reduction, improved quality control, and increased work safety. This work brings data-driven techniques for two predictive maintenance tasks: anomaly detection and event prediction, applied in the real-world use case of a cold forming manufacturing line for consumer lifestyle products by using acoustic emissions sensors in proximity of the dies of the press module. The proposed models are robust and able to cope with problems such as noise, missing values, and irregular sampling. The detected anomalies are investigated by experts and confirmed to correspond to deviations in the normal operation of the machine. Moreover, we are able to find patterns which are related to the events of interest.

Non-destructive and rapid detection of the internal chemical composition of granules samples by spectral transfer

Near-infrared spectroscopy (NIRS) analysis has been widely used in various industries, but the stability and accuracy of the final analysis results rely on the sample uniformity, limiting the application of the technique in the analysis of granules samples. In this work, the protein and amylose of rice as research objects, and proposed to transform the spectrum of the granules into the spectrum of its corresponding powder through spectral transfer based on direct standardization (DS), successfully, followed by using the powder model to predict the new spectrum. After the spectral transfer, the root mean square of validation (RMSEV) values of protein were reduced from 0.3293, 0.6890, and 0.6815 to 0.2602, 0.5106, and 0.4436, respectively. and the RMSEV values of amylose were reduced from 0.9587, 1.0601, 2.8952 ​to 0.7094, 0.8317, and 2.4090, respectively. After using key variables for spectral transfer, the RMSEV values of protein were 0.2410, 0.5336 and 0.4605, respectively, while the RMSEV values of amylose were 0.6760, 0.8442 and 2.3392, respectively. These results clearly show that spectral transfer can ameliorate the problem of large error and poor stability of analytical results due to the irregular sample and uneven component distribution, which will promote the NIRS technique to achieve real-time and accurate online supervision in agricultural production.

Discrete Band-Limited Signal Reconstruction From Irregular Samples

Image reconstruction from discrete samples is fundamental in remote sensing. The reconstruction problem is more complicated when the sample locations are irregularly spaced and employ different aperture functions. Further, not all 2-D sampling configurations permit full image reconstruction of band-limited signals. In this article, 2-D signal reconstruction from irregular samples with variable apertures is considered using theory and examples. Exact reconstruction requires that the signal be band-limited and the sampling matrix be invertable. The results are sensitive to the sample locations and noise in the measurements. Illustrative examples are provided using simulation and actual data from the L-band Soil Moisture Active Passive (SMAP) radiometer. The general approach can be employed with other sensors.

A Near Real-Time Method for Forest Change Detection Based on a Structural Time Series Model and the Kalman Filter

The increasing availability of dense time series of earth observation data has incited a growing interest in time series analysis for vegetation monitoring and change detection. Vegetation monitoring algorithms need to deal with several time series characteristics such as seasonality, irregular sampling intervals, and signal artefacts. While common algorithms based on deterministic harmonic regression models account for intra-annual seasonality, inter-annual variations of the seasonal pattern related to shifts in vegetation phenology due to different temperature and rainfall are usually not accounted for. We propose a transition to stochastic modelling and present a near real-time change detection method that combines a structural time series model with the Kalman filter. The model continuously adapts to new observations and allows to better separate phenology-related deviations from vegetation anomalies or land cover changes. The method is tested in a forest change detection application aiming at the assessment of damages caused by storm events and insect calamities. Forest changes are detected based on the cumulative sum control chart (CUSUM) which is used to decide if new observations deviate from model-based forecasts. The performance is evaluated in two test sites, one in Malawi (dry tropical forest) and one in Austria (temperate deciduous, coniferous and mixed forests) based on Sentinel-2 time series. Both forest areas are characterized by a distinct, but temporally varying leaf-off season. The presented change detection method shows overall accuracies above 99%, users’ accuracies of 76.8% to 88.6%, and producers’ accuracies of 68.2% to 80.4% for the forest change stratum (minimum mapping unit: 0.1 ha). Results are based on visually interpreted points derived by stratified random sampling. A further analysis revealed that increasing the time series density by merging data from two Sentinel-2 orbits yields better forest change detection accuracies in comparison to using data from one orbit only. The resulting increase in users’ accuracy amounts to 7.6%. The presented method is capable of near real-time processing and could be used for a variety of automated forest monitoring applications.

Fractal-interpolation of seismic traces using vertical scale factor with residual behavior

The inaccessibility of the terrain causes an irregular spatial seismic sampling, leading to poor reliability in the images of the subsoil generated by the processing algorithms; an accurate reconstruction of regular data is therefore crucial. This paper presents a fractal reconstruction method based on the self-similarity fractal property of the seismic data, on local and global scales. The method interpolates non-collected samples on the basis of the vertical scale factor value, thus making a solution possible to a system of coupled equations. However, tests with synthetic data show errors, with a tendency related to the scale factor. To reduce the imbalance, this approach takes into account the behavior of the discrepancy between the expected and the estimated values. The method applied to synthetic and real data sampled irregularly indicates that the fine structure of the seismic data can be quickly and accurately reconstructed using the localized fractal approach. The interpolation method can be fruitful in other fields where data need to be reconstructed from data sets not suitably sampled.

An Adaptive Optimal Interpolation Based on Analog Forecasting: Application to SSH in the Gulf of Mexico

AbstractBecause of the irregular sampling pattern of raw altimeter data, many oceanographic applications rely on information from sea surface height (SSH) products gridded on regular grids where gaps have been filled with interpolation. Today, the operational SSH products are created using the simple, but robust, optimal interpolation (OI) method. If well tuned, the OI becomes computationally cheap and provides accurate results at low resolution. However, OI is not adapted to produce high-resolution and high-frequency maps of SSH. To improve the interpolation of SSH satellite observations, a data-driven approach (i.e., constructing a dynamical forecast model from the data) was recently proposed: analog data assimilation (AnDA). AnDA adaptively chooses analog situations from a catalog of SSH scenes—originating from numerical simulations or a large database of observations—which allow the temporal propagation of physical features at different scales, while each observation is assimilated. In this article, we review the AnDA and OI algorithms and compare their skills in numerical experiments. The experiments are observing system simulation experiments (OSSE) on the Lorenz-63 system and on an SSH reconstruction problem in the Gulf of Mexico. The results show that AnDA, with no necessary tuning, produces comparable reconstructions as does OI with tuned parameters. Moreover, AnDA manages to reconstruct the signals at higher frequencies than OI. Finally, an important additional feature for any interpolation method is to be able to assess the quality of its reconstruction. This study shows that the standard deviation estimated by AnDA is flow dependent, hence more informative on the reconstruction quality, than the one estimated by OI.

Occupation density estimation for noisy high-frequency data

This paper studies the nonparametric estimation of occupation densities for semimartingale processes observed with noise. As leading examples we consider the stochastic volatility of a latent efficient price process, the volatility of the latent noise that separates the efficient price from the actually observed price, and nonlinear transformations of these processes. Our estimation methods are decidedly nonparametric and consist of two steps: the estimation of the spot price and noise volatility processes based on pre-averaging techniques and in-fill asymptotic arguments, followed by a kernel-type estimation of the occupation densities. Our spot volatility estimates attain the optimal rate of convergence, and are robust to leverage effects, price and volatility jumps, general forms of serial dependence in the noise, and random irregular sampling. The convergence rates of our occupation density estimates are directly related to that of the estimated spot volatilities and the smoothness of the true occupation densities. An empirical application involving high-frequency equity data illustrates the usefulness of the new methods in illuminating time-varying risks, market liquidity, and informational asymmetries across time and assets.

A Reconfigurable Random Number Generator Based on the Transient Effects of Ring Oscillators

This brief presents a reconfigurable Random Number Generator (RNG) based on transient effect of ring oscillators. Users can select a method based on the irregular sampling of a regular waveform or on the regular sampling of an irregular waveform to obtain a random bit sequence to be used in different applications, such as lightweight cryptography or high-security communication. The entropy is acquired by exploiting Transient Effect Ring Oscillators (TEROs). The proposed fully-digital RNG structure is firstly implemented on a Zynq-7000 FPGA (Field Programmable Gate Array) without any post-processing method such as the Von Neumann. In addition to the RNG structure, an on-the-line test module based on FIPS 140–2 is also implemented to check the randomness of the produced data statistically in real time. Users can change the statistical test parameters according to their desired security levels. Finally, an ASIC (Application Specific Integrated Circuits) implementation of the proposed RNG is done following the Cadence digital design flow for the TSMC 180 nm CMOS process. The implemented ASIC design occupies an area of 0.85 mm x 0.85 mm and the estimated power required is 11.827 mW.