Reconstruction Error Research Articles

Robust Principal Component Analysis Based on Fuzzy Local Information Reservation.

Principal Component Analysis (PCA) aims to acquire the principal component space containing the essential structure of data, instead of being used for mining and extracting the essential structure of data. In other words, the principal component space contains not only information related to the essential structure of data but also some unrelated information. This frequently occurs when the intrinsic dimensionality of data is unknown or when it has complex distribution characteristics such as multi-modalities, manifolds, etc. Therefore, it is unreasonable to identify noise and useful information based solely on reconstruction error. For this reason, PCA is unsuitable as a preprocessing technique for most applications, especially in noisy environment. To solve this problem, this paper proposes robust PCA based on fuzzy local information reservation (FLIPCA). By analyzing the impact of reconstruction error on sample discriminability, FLIPCA provides a theoretical basis for noise identification and processing. This not only greatly improves its robustness but also extends its applicability and effectiveness as a data preprocessing technique. Meanwhile, FLIPCA maintains consistent mathematical descriptions with traditional PCA while having few adjustable hyperparameters and low algorithmic complexity. Finally, we conducted comprehensive experiments on synthetic and real-world datasets, which substantiated the superiority of our proposed algorithm.

Laplacian Projection Based Global Physical Prior Smoke Reconstruction.

We present a novel framework for reconstructing fluid dynamics in real-life scenarios. Our approach leverages sparse view images and incorporates physical priors across long series of frames, resulting in reconstructed fluids with enhanced physical consistency. Unlike previous methods, we utilize a differentiable fluid simulator (DFS) and a differentiable renderer (DR) to exploit global physical priors, reducing reconstruction errors without the need for manual regularization coefficients. We introduce divergence-free Laplacian eigenfunctions (div-free LE) as velocity bases, improving computational efficiency and memory usage. By employing gradient-related strategies, we achieve better convergence and superior results. Extensive experiments demonstrate the effectiveness of our method, showcasing improved reconstruction quality and computational efficiency compared to existing approaches. We validate our approach using both synthetic and real data, highlighting its practical potential.

Acoustic sensing and autoencoder approach for abnormal gas detection in a spent nuclear fuel canister mock-up

Currently, spent nuclear fuel (SNF) from commercial nuclear power plants is stored in stainless-steel canisters for interim dry storage. To provide an inert environment, these canisters are backfilled with helium after vacuum drying. However, the helium environment may be contaminated during extended storage because of the material degradation. For example, the heavier fission gas xenon may be released from the fuel rods into the canister cavity should the fuel cladding be breached. Other gases such as air and water vapor may also be present as a result of leakage caused by chloride-induced stress corrosion cracking on the canister walls or by insufficient vacuum drying. Therefore, monitoring the gas composition can provide critical information about the health of SNF canisters. In this study, noninvasive testing was conducted on a 2/3-scaled SNF canister mock-up using acoustic sensing. Ultrasonic transducers were placed on the exterior surface of the canister to probe the gas composition. A dataset was collected by sealing the canister mock-up and introducing up to 1.53% argon or 1.29% air into the helium background gas. Three methods were used to detect changes in the gas composition: the time-of-flight (TOF) method, the differential method, and the autoencoder method. Results showed that the TOF method had sufficient resolution to detect abnormal gas concentrations of less than 1.0%. The differential method demonstrated a periodic in-phase and out-of-phase behavior between the benchmark (i.e., pure helium) and abnormal (i.e., with argon or air) state signals. The variational autoencoder (VAE) and the Wasserstein autoencoder (WAE) were trained on the benchmark data and were applied directly to the abnormal state data. It was found that both the unsupervised VAE and the WAE were able to distinguish the benchmark and abnormal states of the canister mock-up based on the reconstruction error.

Radar False Target Recognition Method Based on Reconstruction Error Distribution Characteristics of Frequency Response From Multi-CPIs

Radar False Target Recognition Method Based on Reconstruction Error Distribution Characteristics of Frequency Response From Multi-CPIs

Height reconstructions from geometric wavefronts using vision ray metrology

A recently reported vision ray metrology technique [Opt. Express 29, 43480 (2021)OPEXFF1094-408710.1364/OE.443550] measures geometric wavefronts with high precision. This paper introduces a method to convert these wavefront data into height information, focusing on the impact of back surface flatness and telecentricity errors on measurement accuracy. Systematic errors from these factors significantly affect height measurements. Using ray trace simulations, we estimate reconstruction errors with various plano-concave and plano-convex elements. We also developed a calibration technique to mitigate telecentricity errors, achieving submicron accuracy in surface reconstruction. This study provides practical insights into vision ray metrology systems, highlighting validity limits, emphasizing the importance of calibration for larger samples, and establishing system alignment tolerances. The reported technique for the conversion of geometric wavefronts to surface topography employs a direct non-iterative ray-tracing-free method. It is ideally suited for reference-free metrology with application to freeform optics manufacturing.

Compact speckle spectrometer based on CNN-LSTM denoising.

We propose a compact speckle spectrometer that utilizes micro-nanostructures processed by femtosecond lasers on sapphire surfaces as scattering media. The spectral resolution is 0.5 nm, and the entire system is compact and stable. At the same time, the convolutional long short-term memory network (CNN-LSTM) was introduced into the denoising algorithm. Compared with traditional reconstruction algorithms, this method not only ensures rapid spectral reconstruction but also offers better reconstruction accuracy. It can effectively reduce the reconstruction error caused by the reduction of speckle autocorrelation caused by environmental noise and prolong the stability time of the system.

MSDG: Multi-Scale Dynamic Graph Neural Network for Industrial Time Series Anomaly Detection

A large number of sensors are typically installed in industrial plants to collect real-time operational data. These sensors monitor data with time series correlation and spatial correlation over time. In previous studies, GNN has built many successful models to deal with time series data, but most of these models have fixed perspectives and struggle to capture the dynamic correlations in time and space simultaneously. Therefore, this paper constructs a multi-scale dynamic graph neural network (MSDG) for anomaly detection in industrial sensor data. First, a multi-scale sliding window mechanism is proposed to input different scale sensor data into the corresponding network. Then, a dynamic graph neural network is constructed to capture the spatial–temporal dependencies of multivariate sensor data. Finally, the model comprehensively considers the extracted features for sequence reconstruction and utilizes the reconstruction errors for anomaly detection. Experiments have been conducted on three real public datasets, and the results show that the proposed method outperforms the mainstream methods.

Mark3D - A semi-automated open-source toolbox for 3D head- surface reconstruction and electrode position registration using a smartphone camera video.

Source localization in EEG necessitates co-registering the EEG sensor locations with the subject's MRI, where EEG sensor locations are typically captured using electromagnetic tracking or 3D scanning of the subject's head with EEG cap, using commercially available 3D scanners. Both methods have drawbacks, where, electromagnetic tracking is slow and immobile, while 3D scanners are expensive. Photogrammetry offers a cost-effective alternative but requires multiple photos to sample the head, with good spatial sampling to adequately reconstruct the head surface. Post-reconstruction, the existing tools for electrode position labelling on the 3D head-surface have limited visual feedback and do not easily accommodate customized montages, which are typical in multi-modal measurements. We introduce Mark3D, an open-source, integrated tool for 3D head-surface reconstruction from phone camera video. It eliminates the need for keeping track of spatial sampling during image capture for video-based photogrammetry reconstruction. It also includes blur detection algorithms, a user-friendly interface for electrode and tracking, and integrates with popular toolboxes such as FieldTrip and MNE Python. The accuracy of the proposed method was benchmarked with the head-surface derived from a commercially available handheld 3D scanner Einscan-Pro + (Shining 3D Inc.,) which we treat as the "ground truth". We used reconstructed head-surfaces of ground truth (G1) and phone camera video (M1080) to mark the EEG electrode locations in 3D space using a dedicated UI provided in the tool. The electrode locations were then used to form pseudo-specific MRI templates for individual subjects to reconstruct source information. Somatosensory source activations in response to vibrotactile stimuli were estimated and compared between G1 and M1080. The mean positional errors of the EEG electrodes between G1 and M1080 in 3D space were found to be 0.09 ± 0.01mm across different cortical areas, with temporal and occipital areas registering a relatively higher error than other regions such as frontal, central or parietal areas. The error in source reconstruction was found to be 0.033 ± 0.016mm and 0.037 ± 0.017mm in the left and right cortical hemispheres respectively.

Robust block tensor PCA with F-norm projection framework

Tensor principal component analysis (TPCA), also known as Tucker decomposition, ensures that the extracted “core tensor” maximizes the variance of the sample projections. Nevertheless, this method is particularly susceptible to noise and outliers. This is due to the utilization of the squared F-norm as the distance metric. In addition, it lacks constraints on the discrepancies between the original tensors and the projected tensors. To address these issues, a novel tensor-based trigonometric projection framework is proposed using F-norm to measure projection distances. Tensor data are first processed utilizing a blocking recombination technique prior to projection, thus enhancing the representation of the data at a local spatio-temporal level. Then, we present a block TPCA with the F-norm metric (BTPCA-F) and develop an iterative greedy algorithm for solving BTPCA-F. Subsequently, regarding the F-norm projection relation as the “Pythagorean Theorem”, we provide three different objective functions, namely, the tangent, cosine and sine models. These three functions directly or indirectly achieve the two objectives of maximizing projection distances and minimizing reconstruction errors. The corresponding tangent, cosine and sine solution algorithms based on BTPCA-F (called tan-BTPCA-F, cos-BTPCA-F and sin-BTPCA-F) are presented to optimize the objective functions, respectively. The convergence and rotation invariance of these algorithms are rigorously proved theoretically and discussed in detail. Lastly, extensive experimental results illustrate that the proposed methods significantly outperform the existing TPCA and the related 2DPCA algorithms.

Large-Kernel Central Block Masked Convolution and Channel Attention-Based Reconstruction Network for Anomaly Detection of High-Resolution Hyperspectral Imagery

In recent years, the rapid advancement of drone technology has led to an increasing use of drones equipped with hyperspectral sensors for ground imaging. Hyperspectral data captured via drones offer significantly higher spatial resolution, but this also introduces more complex background details and larger target scales in high-resolution hyperspectral imagery (HRHSI), posing substantial challenges for hyperspectral anomaly detection (HAD). Mainstream reconstruction-based deep learning methods predominantly emphasize spatial local information in hyperspectral images (HSIs), relying on small spatial neighborhoods for reconstruction. As a result, large anomalous targets and background details are often well reconstructed, leading to poor anomaly detection performance, as these targets are not sufficiently distinguished from the background. To address these limitations, we propose a novel HAD network for HRHSI based on large-kernel central block masked convolution and channel attention, termed LKCMCA. Specifically, we first employ the pixel-shuffle technique to reduce the size of anomalous targets without losing image information. Next, we design a large-kernel central block masked convolution to make the network pay more attention to the surrounding background information, enabling better fusion of the information between adjacent bands. This, coupled with an efficient channel attention mechanism, allows the network to capture deeper spectral features, enhancing the reconstruction of the background while suppressing anomalous targets. Furthermore, we introduce an adaptive loss function by down-weighting anomalous pixels based on the mean absolute error. This loss function is specifically designed to suppress the reconstruction of potentially anomalous pixels during network training, allowing our model to be considered an excellent background reconstruction network. By leveraging reconstruction error, the model effectively highlights anomalous targets. Meanwhile, we produced four benchmark datasets specifically for HAD tasks using existing HRHSI data, addressing the current shortage of HRHSI datasets in the HAD field. Extensive experiments demonstrate that our LKCMCA method achieves superior detection performance, outperforming ten state-of-the-art HAD methods on all datasets.

Capturing forceful interaction with deformable objects using a deep learning-powered stretchable tactile array

Capturing forceful interaction with deformable objects during manipulation benefits applications like virtual reality, telemedicine, and robotics. Replicating full hand-object states with complete geometry is challenging because of the occluded object deformations. Here, we report a visual-tactile recording and tracking system for manipulation featuring a stretchable tactile glove with 1152 force-sensing channels and a visual-tactile joint learning framework to estimate dynamic hand-object states during manipulation. To overcome the strain interference caused by contact with deformable objects, an active suppression method based on symmetric response detection and adaptive calibration is proposed and achieves 97.6% accuracy in force measurement, contributing to an improvement of 45.3%. The learning framework processes the visual-tactile sequence and reconstructs hand-object states. We experiment on 24 objects from 6 categories including both deformable and rigid ones with an average reconstruction error of 1.8 cm for all sequences, demonstrating a universal ability to replicate human knowledge in manipulating objects with varying degrees of deformability.

A D* orthogonal matching pursuit algorithm for time-varying channel estimation.

Orthogonal matching pursuit (OMP) combined with the A* search algorithm (A*OMP) exhibits robust reconstruction capabilities for synthesizing sparse data and signals, achieving relatively low reconstruction errors and a higher exact recovery ability than conventional OMP. However, A*OMP is only suitable for static channel estimation and cannot be applied to dynamic scenarios. This is because the channel delays for several consecutive orthogonal frequency-division multiplexing blocks per frame are similar and the path gains exhibit temporal correlation. This paper introduces a dynamic OMP approach (D*OMP) that employs a heuristic function of A*OMP and a unique reverse process, enabling sparse solutions to be identified in unknown and changing environments. The proposed method is highly practical for joint channel estimation across multiple blocks. Simulation and sea trial results indicate that D*OMP not only possesses superior channel recovery accuracy, but also has a more efficient channel reconstruction process, outperforming both A*OMP and conventional OMP.

The Perturbation Analysis of Nonconvex Low-Rank Matrix Robust Recovery.

In this article, we bring forward a completely perturbed nonconvex Schatten p -minimization to address a model of completely perturbed low-rank matrix recovery (LRMR). This article based on the restricted isometry property (RIP) and the Schatten- p null space property (NSP) generalizes the investigation to a complete perturbation model thinking over not only noise but also perturbation, and it gives the RIP condition and the Schatten- p NSP assumption that guarantee the recovery of low-rank matrix and the corresponding reconstruction error bounds. In particular, the analysis of the result reveals that in the case that p decreases 0 and for the complete perturbation and low-rank matrix, the condition is the optimal sufficient condition (Recht et al., 2010). In addition, we study the connection between RIP and Schatten- p NSP and discern that Schatten- p NSP can be inferred from the RIP. The numerical experiments are conducted to show better performance and provide outperformance of the nonconvex Schatten p -minimization method comparing with the convex nuclear norm minimization approach in the completely perturbed scenario.

Detection of Cyber-Physical Attacks in Additive Manufacturing: An LSTM-Based Autoencoder Method Utilizing Reconstruction Error Analysis from Side-Channel Monitoring

To identify cyber-physical threats in additive manufacturing systems, this study proposes an advanced technique utilizing data from side-channel monitoring. The method combines several key approaches for preprocessing, analyzing, and classifying time-series data, ensuring robust attack detection capabilities. A predefined Window Sliding (FWS) preprocessing method segments continuous time-series data into manageable windows of specified size, making analysis more efficient. Next, we employ Discrete Wavelet Transform (DWT) to extract features from each window, capturing essential information across various frequency bands. Particle Swarm Optimization (PSO) is then used to refine the DWT coefficients, isolating the most valuable features to enhance classification performance by focusing on the most informative characteristics. The optimal feature set is used to train a deep learning (DL) model capable of identifying anomalies through reconstruction errors, specifically an LSTM-A autoencoder. Our results demonstrate that this approach can distinguish between normal and attack scenarios with an accuracy rate of 99% when applied to side-channel attack detection in additive manufacturing. This method provides a scalable and adaptable solution to safeguard cyber-physical systems against sophisticated cyberattacks while improving detection accuracy.

Hierarchical Mixed-Precision Post-Training Quantization for SAR Ship Detection Networks

Convolutional neural network (CNN)-based synthetic aperture radar (SAR) ship detection models operating directly on satellites can reduce transmission latency and improve real-time surveillance capabilities. However, limited satellite platform resources present a significant challenge. Post-training quantization (PTQ) provides an efficient method for pre-training neural networks to effectively reduce memory and computational resources without retraining. Despite this, PTQ faces the challenge of maintaining model accuracy, especially at low-bit quantization (e.g., 4-bit or 2-bit). To address this challenge, we propose a hierarchical mixed-precision post-training quantization (HMPTQ) method for SAR ship detection neural networks to reduce quantization error. This method encompasses a layerwise precision configuration based on reconstruction error and an intra-layer mixed-precision quantization strategy. Specifically, our approach initially utilizes the activation reconstruction error of each layer to gauge the sensitivity necessary for bit allocation, considering the interdependencies among layers, which effectively reduces the complexity of computational sensitivity and achieves more precise quantization allocation. Subsequently, to minimize the quantization error of the layers, an intra-layer mixed-precision quantization strategy based on probability density assigns a greater number of quantization bits to regions where the probability density is low for higher values. Our evaluation on the SSDD, HRSID, and LS-SSDD-v1.0 SAR Ship datasets, using different detection CNN models, shows that the YOLOV9c model with mixed-precision quantization at 4-bit and 2-bit for weights and activations achieves only a 0.28% accuracy loss on the SSDD dataset, while reducing the model size by approximately 80%. Compared to state-of-the-art methods, our approach maintains competitive accuracy, confirming the superior performance of the HMPTQ method over existing quantization techniques.

Inclination information-aided circumferential deformation reconstruction error correction method for CFRP belt integrated with optical fiber sensor

In view of the problem that the distributed optical fiber-based deformation reconstruction method can produce serious accumulated errors, this paper developed a smart fiber optic sensor belt that employs a carbon-fiber composite material as the matrix and strain-sensing fiber optics as the sensing element and is based on the key-point dip angle information acquired by a high-precision tilt sensor. The method of in situ correction for the cumulative error of circumferential deformation was investigated, and the particle swarm optimization algorithm was adopted to perform in situ correction for the reconstructed circumferential deformation field of the sensor belt to reduce the cumulative measurement error of the circumferential deformation field. Simulations and indoor tests were systematically performed. After the correction, the errors of the simulation and indoor tests were reduced from 210.83 and 235.82 mm to 2.17 and 4.04 mm, respectively, which verified the effectiveness and accuracy of the aforementioned method.

The Parameterized Oceanic Front-Guided PIX2PIX Model: A Limited Data-Driven Approach to Oceanic Front Sound Speed Reconstruction

In response to the demand for high-precision acoustic support under the condition of limited data, this study utilized high-resolution reanalysis data and in situ observation data to extract the Kuroshio Extension Front (KEF) section through front-line identification methods. By combining the parameterized oceanic front model and the statistical features of big data, the parameterized oceanic front was reconstructed. A proxy dataset was generated using the Latin hypercube sampling method, and the sound speed reconstruction model based on the PIX2PIX model was trained and validated using single sound speed profiles at different positions of the oceanic front, combined with the parameterized oceanic front model. The experimental results show that the proposed sound speed reconstruction model can significantly improve the reconstruction accuracy by introducing the parameterized front model as an additional input, especially in the shallow-water area. The mean absolute error (MAE) of the full-depth sound speed reconstruction for this model is 0.63~0.95 m·s−1, and the structural similarity index (SSIM) is 0.76~0.78. The MAE of the sound speed section within a 1000 m depth is reduced by 6.50~37.62%, reaching 1.95~3.31 m·s−1. In addition, the acoustic support capabilities and generalization of the model were verified through ray tracing models and in situ data. This study contributes to advancing high-precision acoustic support in data-limited oceanic environments, laying a solid groundwork for future innovations in marine acoustics.

Clustering dynamic networks by discriminating roles of vertices and capturing temporality with subsequent feature projection

Clustering dynamic networks has gained popularity due to the need to analyze complex systems that evolve over time, which cannot be fully characterized by traditional static models. It is highly non-trivial in comparison to clustering static network since it requires simultaneously to balance clustering accuracy and clustering drift, where clustering accuracy measures how clustering reflects structure of graph at current time, and clustering drift quantifies how clustering smoothes historical snapshot(s). In this study, we propose an algorithm clustering dynamic network by discriminating roles of vertices and capturing temporality with subsequent feature projection (CDN-DRCT). Specifically, clustering accuracy is achieved by factorizing high-order matrix of slice at current time, and vertices are divided into static and dynamic ones by the reconstruction errors. Finally, the proposed algorithm measures temporality of networks with a projection matrix, which connects subsequent features at the previous and current time, thereby enhancing clustering drift of clusters. In this case, temporality of dynamic networks is characterized from vertex and global level, providing a better way to balance clustering accuracy and clustering drift. Experimental results on 10 typical dynamic networks demonstrate the proposed algorithm is superior to baselines in terms of accuracy as well efficiency.

Tool wear prediction based on K-means and Adaboost auto-encoder

Abstract A new tool wear prediction model is proposed to address the tool wear issue, aimed at monitoring tool wear based on specific task requirements and guiding tool replacement during actual cutting operations. In the data preprocessing phase, tool wear states are classified using unsupervised K-means clustering. The time, frequency, and time-frequency domain features are then labeled and fused using an autoencoder neural network applied to the original set of signal features from the tool. For tool wear prediction, an enhanced autoencoder neural network leveraging AdaBoost is employed to establish the prediction model. The reconstruction error serves as the chosen loss function to assess the autoencoder's performance, taking into account data correlation and the inherent lossy nature of the autoencoder. Experimental results from real machining data obtained from a CNC milling machine demonstrate that the proposed model achieves higher prediction accuracy while reducing data dimensions.

The Validity of a Smartphone-Based Method for Acquiring 3D Images of the Face.

Objectives. To evaluate the accuracy and reproducibility of measurements obtained using the Bellus3D Face Application on a mobile smartphone by comparing them to direct measurements on pre-marked and blank face scans. Materials and Methods. Twenty-five healthy young adults (six males and nineteen females; age range 20-30 years) were included in this prospective cross-sectional study, with the only exclusion criterion being the presence of significant facial hair interfering with the placement and visualization of landmarks. Image acquisitions were performed using an iPhone XR with the Bellus3D FaceApp face scanning application, an iOS application for smartphones. Ten single midfacial and five paired bilateral landmarks were defined and marked. Two face scans were performed on each patient, both on blank and marked faces, and distances were measured directly with calipers and digitally. Results. The random error values were 1.0 mm and 0.4 mm for the manual point placement and measurements and virtual point placement on blank faces, respectively. The two methods used (the direct method and acquisition on faces with landmarks) demonstrate relatively similar reliability (ICC > 0.8); however, a paired t-test showed that the differences between several measurements were statistically significant (p < 0.05). Regardless of the method used, there was a systematic error for various values that included the nose and mouth (p < 0.05). The measurements demonstrating the most significant differences between the methods were those that included the tip of the nose, with the mean differences being -4.4-3.3 mm. The measurements of the distances that estimate face "depth" showed the greatest consistency irrespective of the tested method (p > 0.05 and ICC > 0.8). Conclusions. The use of the Bellus3D FaceApp is precise and reproducible for certain areas of the face, but digital reconstruction errors prohibit, for the time being, the use of this technology in everyday clinical practice. The noted discrepancies were consistent and more prevalent for specific areas such as the tip of the nose. Further investigations are required to determine other sources of error and for other smartphone-based applications released for 3D face image acquisitions.