Input Samples Research Articles

MK-BMC: a Multi-Kernel framework with Boosted distance metrics for Microbiome data for Classification.

Research on human microbiome has suggested associations with human health, opening opportunities to predict health outcomes using microbiome. Studies have also suggested that diverse forms of taxa such as rare taxa that are evolutionally related and abundant taxa that are evolutionally unrelated could be associated with or predictive of a health outcome. Although prediction models were developed for microbiome data, no prediction models currently exist that use multiple forms of microbiome-outcome associations. We developed MK-BMC, a Multi-Kernel framework with Boosted distance Metrics for Classification using microbiome data. We propose to first boost widely used distance metrics for microbiome data using taxon-level association signal strengths to up-weight taxa that are potentially associated with an outcome of interest. We then propose a multi-kernel prediction model with one kernel capturing one form of association between taxa and the outcome, where a kernel measures similarities of microbiome compositions between pairs of samples being transformed from a proposed boosted distance metric. We demonstrated superior prediction performance of (i) boosted distance metrics for microbiome data over original ones and (ii) MK-BMC over competing methods through extensive simulations. We applied MK-BMC to predict thyroid, obesity, and inflammatory bowel disease status using gut microbiome data from the American Gut Project and observed much-improved prediction performance over that of competing methods. The learned kernel weights help us understand contributions of individual microbiome signal forms nicely. Source code together with a sample input dataset is available at https://github.com/HXu06/MK-BMC.

An Interpretable Deep Embedding Model for Few and Imbalanced Biomedical Data.

In healthcare, training examples are usually hard to obtain (e.g., cases of a rare disease), or the cost of labelling data is high. With a large number of features ( p) be measured in a relatively small number of samples ( N), the "big p, small N" problem is an important subject in healthcare studies, especially on the genomic data. Another major challenge of effectively analyzing medical data is the skewed class distribution caused by the imbalance between different class labels. In addition, feature importance and interpretability play a crucial role in the success of solving medical problems. Therefore, in this paper, we present an interpretable deep embedding model (IDEM) to classify new data having seen only a few training examples with highly skewed class distribution. IDEM model consists of a feature attention layer to learn the informative features, a feature embedding layer to directly deal with both numerical and categorical features, a siamese network with contrastive loss to compare the similarity between learned embeddings of two input samples. Experiments on both synthetic data and real-world medical data demonstrate that our IDEM model has better generalization power than conventional approaches with few and imbalanced training medical samples, and it is able to identify which features contribute to the classifier in distinguishing case and control.

Rejecting Unknown Gestures Based on Surface-Electromyography Using Variational Autoencoder.

The conventional surface electromyography (sEMG)-based gesture recognition systems exhibit impressive performance in controlled laboratory settings. As most systems are trained in a closed-set setting, the systems's performance may see significant deterioration when novel gestures are presented as imposter. In addition, the state-of-the-art generative and discriminative methods have achieved considerable performance on high-density sEMG signals. This can be seen as an unrealistic setting as the real-world muscle computer interface are mainly comprised of sparse multichannel sEMG signals. In this work, we propose a novel variational autoencoder based approach for open-set gesture recognition based on sparse multichannel sEMG signals. Using the predefined corresponding latent conditional distribution of known gestures, the conditional Gaussian distribution of each known gesture is learned. Those samples with low probability density are identified as unknown gestures. The sEMG signals of known gestures are classified using the Kullback-Leibler divergences between the predefined prior distributions and input samples. The proposed approach is evaluated using three benchmark sparse multichannel sEMG databases. The experimental results demonstrate that our approach outperforms the existing open-set sEMG-based gesture recognition approaches and achieves a better trade-off between classifying known gestures and rejecting unknown gestures.

A Novel Approach for Single-Shot Target Recognition by Matching Graph of Feature Pixels

Image feature extraction and matching constitute fundamental steps in computer vision and image analysis, enabling various applications ranging from object recognition to scene reconstruction. Object recognition is one of the major applications of artificial intelligence. Since last two decades, different techniques for object recognition were given by researchers from across the world. This paper presents a novel approach for automatic target recognition in autonomous weapons by matching connected graph of feature pixels. The Moving and Stationary Target Acquisition and Recognition (MSTAR) is taken for experimental purpose. The Proposed model has three stages for a target recognition; Image Denoising, Feature extraction & graph generation, and feature matching. Firstly, a sample image of targets taken from the MSTAR is taken as an input. Further, input sample is de-noised and features are extracted using Speeded-Up Robust Features (SURF) algorithm. Thereafter, a graph is created using feature pixels and this complete process is repeated with test image. Further, both graphs are compared using a proposed feature graph matching algorithm which results an recognition accuracy score. Proposed model is tested over the MSTAR dataset and it resulted in varying accuracy at different feature matching thresholds.

Separation of platelets by size in a microfluidic device based on controlled incremental filtration.

The significant biological and functional differences between small and large platelets suggested by recent studies could have profound implications for transfusion medicine. However, investigating the relationship between platelet size and function is challenging because separating platelets by size without affecting their properties is difficult. A standard approach is centrifugation, but it inevitably leads to premature activation and aggregation of separated platelets. This paper describes the development and validation of a microfluidic device based on controlled incremental filtration (CIF) for separating platelets by size without the cell damage and usability limitations associated with centrifugation. Platelet samples derived from whole blood were used to evaluate the dependence of the CIF device separation performance on design parameters and flow rate, and to compare the properties of PLT fractions generated by the CIF device with those produced using a centrifugation protocol in a split-sample study. This was accomplished by quantifying the platelet size distribution, mean platelet volume (MPV), platelet-large cell ratio (P-LCR) and platelet activation before and after processing for all input and output samples. The 'large platelet' fractions produced by the CIF device and the centrifugation protocol were essentially equivalent (no significant difference in MPV and P-LCR). Platelets in the 'small platelet' fraction produced by the CIF device were significantly smaller than those produced by centrifugation (lower MPV and P-LCR). This was because the CIF 'small platelet' fraction was contaminated by much fewer large platelets (∼2-times lower recovery of >12 fL platelets) and retained the smallest platelets that were discarded by the centrifugation protocol. There was no significant difference in platelet activation between the two methods. However, centrifugation required a substantial amount of additional anticoagulant to prevent platelet aggregation during pelleting. Unlike centrifugation, the CIF device offered continuous, flow-through, single-step processing that did not cause platelet aggregation. Such a capability has the potential to accelerate the basic studies of the relationship between platelet size and function, and ultimately improve transfusion practice, particularly in the pediatric setting, where the need for low-volume, high-quality platelet transfusions is most urgent.

Estimating the temporal scale of lagged responses in species abundance and occurrence

AbstractUnderstanding drivers of species distributions is a central theme in ecology, and species are known to respond to drivers at multiple spatiotemporal scales. Yet, little attention has been paid to the temporal dimensions of predictors of species abundance or occurrence. Studies considering lagged responses typically contrast few points in time, rather than integrating multiple temporal scales. Here, I transfer an approach developed for the integration of predictors across spatial scales into the temporal realm. In this approach, a weighted mean predictor is used in a generalized linear model for abundance or occurrence; weights are estimated for each temporal scale based on a parametric function, and the parameters defining the function are estimated within the model. In a simulation study using a half‐normal declining weight function, I show that the scale parameter σ (measuring temporal scale of response) and effect strength of the predictor, , can be estimated accurately in abundance models with and without imperfect detection for a range of sample sizes and input values of σ and . Accuracy for both parameters was lower in occurrence‐based models. Estimates of σ improved in accuracy with increasing , and precision of both estimates tended to be lower in models with imperfect detection. A higher resolution time series of predictor values improved estimates of both parameters, and lower detection probability most strongly increased bias in in occurrence models. The approach can be viewed as a more parsimonious but less flexible case of memory models or distributed lag models. It is easily incorporated into standard methods used to model animal abundance and occurrence and therefore hopefully will open up avenues for future research into lagged responses in wildlife distribution to environmental conditions.

Chloride-induced concrete deterioration monitoring using advanced ultrasonic pulse wave analysis based on convolutional neural network

This research explores the potential of deep learning techniques, specifically the convolutional neural network (CNN) architecture, for classifying concrete crack levels based on an acceptable threshold of concrete cracking. The classification model utilizes ultrasonic pulse wave data collected from concrete cube specimens before and after undergoing an accelerated corrosion process. A total of 108 concrete specimens, representing three different mix designs, three corrosion levels, and four concrete cover thicknesses, were utilized in this study. The collected data was employed to train CNN models, specifically leveraging the GoogLeNet and SqueezeNet architectures. Various input sampling rates, input lengths, and hyperparameters were explored to determine the optimal training setup, yielding the best prediction performance. The results demonstrate that the optimized models achieve an 84% accuracy in distinguishing cracks below and above the acceptable threshold. Therefore, it can be concluded that the CNN method holds potential for in-situ sensors aimed at monitoring chloride-induced deterioration in concrete structures.

Learning-Based Adaptive Optimal Control of Linear Time-Delay Systems: A Policy Iteration Approach

This paper studies the adaptive optimal control problem for a class of linear time-delay systems described by delay differential equations (DDEs). A crucial strategy is to take advantage of recent developments in reinforcement learning (RL) and adaptive dynamic programming (ADP) and develop novel methods to learn adaptive optimal controllers from finite samples of input and state data. In this paper, the data-driven policy iteration (PI) is proposed to solve the infinite-dimensional algebraic Riccati equation (ARE) iteratively in the absence of exact model knowledge. Interestingly, the proposed recursive PI algorithm is new in the present context of continuous-time time-delay systems, even when the model knowledge is assumed known. The efficacy of the proposed learning-based control methods is validated by means of practical applications arising from metal cutting and autonomous driving.

Text Localization and Enhancement of Mobile Camera based Complex Natural Bilingual Text Scene Images

In this paper, we proposed a text localization and enhancement of the Mobile camera based complex natural scene images using fixed coordinates for removal of unwanted non text regions and making it localized. The morphological erosion technique is applied for the smoothness of foreground text regions. Here, the experimental working nature of the proposed method is first localized color input image is converted into a gray level, then a median filter is applied for the removal of noises. After that, the input image is further converted into binary image. In addition to this, morphological dilation and erosion operations also applied for connecting the broken pixels. Further, the connected component segmentation technique is applied to extract the foreground text regions from the binary input image. The real time input samples are collected by using mobile camera from the various categories of natural scenes of monolingual/bilingual text such as Stone, Wall, Poster and Iron-based. A total of 140 real time samples of natural scene text images are considered for the experimental setup. The experimental results shown that 132 real time sample images are localized and then obtained the accuracy of 94.24%. Further, bilingual text regions are enhanced using hybrid approach and obtain the segmentation accuracy of 83.17%. The novelty of the paper is working on the real time data set and it contains a mixture of printed and handwritten natural scene text images.

Cross-Anatomy Transfer Learning via Shape-Aware Adaptive Fine-Tuning for 3D Vessel Segmentation.

Deep learning methods have recently achieved remarkable performance in vessel segmentation applications, yet require numerous labor-intensive labeled data. To alleviate the requirement of manual annotation, transfer learning methods can potentially be used to acquire the related knowledge of tubular structures from public large-scale labeled vessel datasets for target vessel segmentation in other anatomic sites of the human body. However, the cross-anatomy domain shift is a challenging task due to the formidable discrepancy among various vessel structures in different anatomies, resulting in the limited performance of transfer learning. Therefore, we propose a cross-anatomy transfer learning framework for 3D vessel segmentation, which first generates a pre-trained model on a public hepatic vessel dataset and then adaptively fine-tunes our target segmentation network initialized from the model for segmentation of other anatomic vessels. In the framework, the adaptive fine-tuning strategy is presented to dynamically decide on the frozen or fine-tuned filters of the target network for each input sample with a proxy network. Moreover, we develop a Gaussian-based signed distance map that explicitly encodes vessel-specific shape context. The prediction of the map is added as an auxiliary task in the segmentation network to capture geometry-aware knowledge in the fine-tuning. We demonstrate the effectiveness of our method through extensive experiments on two small-scale datasets of coronary artery and brain vessel. The results indicate the proposed method effectively overcomes the discrepancy of cross-anatomy domain shift to achieve accurate vessel segmentation for these two datasets.

Bilateral‐Branch Signal Feature Enhancement Network for RFF Data Enhancement

The clarity of radio frequency fingerprint (RFF) features is an important factor affecting the recognition effect of the classifier. Recognition of RFF signals is essential in radio frequency (RF) perception. However, the RF signal is easily affected by interference and noise during transmission, which makes the distribution boundary of the characteristics of the RFF unclear. It then affects the recognition effect of the classifier. The common preprocessing algorithms, such as denoising and channel compensation, can improve the performance of communication demodulation but seriously affect the unique signal distortion of RFF characteristics. The spectral overlap between the useful signal and the noise is serious when the spectrum is broad, so it is difficult to separate the signal from the noise by traditional methods effectively. Therefore, this paper proposes a data enhancement technology for RF signals named random railings to increase the diversity of input samples and improve the overall performance of the model. At the same time, this paper also proposes a bilateral‐branch signal feature enhancement network. The dual branch structure of BBSEN extracts the features of the enhanced time‐domain signal and the spectrum obtained by the short‐time Fourier transform (STFT) of the signal and then achieves feature fusion through the improved self‐attention mechanism to obtain the RFF features, which are more suitable for the interference environment. Experiments show that the algorithm proposed in this paper can effectively improve the antijamming environment of RF signals and even can be used after preprocessing algorithms such as denoising and further improve the recognition rate of high signal‐to‐noise ratio signals.

Low-Light Enhancement Method Based on a Retinex Model for Structure Preservation

Enhancing low-light image visibility is a critical task in computer vision since it helps to improve input for high-level algorithms. High-quality images typically have clear structural information. In previous studies, due to the lack of proper structural guidance, restored images had some problems, such as unclear structural areas and overexposed or underexposed local areas. To address the above problems, in this paper, we introduce a coefficient of variation (COV) with excellent performance in maintaining structural information, and then we propose a low-light image enhancement method that utilizes the COV to extract structural information from images. First, we apply a traditional retinex model to estimate both reflectance and illumination. Second, we use the COV to indicate the degree of dispersion of the input sample, which enables us to obtain a robust structure-distinguishing weight map for low-light images. The weight map is adaptively divided to obtain a structural weight map, which is then used to enhance the gradient image. This process is applied before the reflectance layer of the retinex model. Finally, the result is obtained by using the block coordinate descent method. According to extensive experiments, outstanding results can be achieved by our proposed method in terms of both subjective and objective evaluation metrics in comparison with other state-of-the-art methods. The source code is available at our website <xref ref-type="fn" rid="fn1" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">¹</xref> <fn id="fn1" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><label>¹</label> <uri>https://github.com/bbxavi/spcv22</uri> </fn> .

Recognizing Multitype Misalignments in Wireless EV Chargers With Orientation-Sensitive Coils: A Data-Driven Strategy Using Improved ResNet

In wireless electric vehicle charging systems, misalignments between ground-assembled transmitter and vehicle-assembled receiver are undesirable but inevitable, which may cause efficiency deterioration and even thermal risks. The recognition of horizontal misalignment, including that in the longitudinal and lateral direction, has been made gradual progress by previous works. While yaw angle, with rare attention, can also degrade the charging performance, especially for those employing non-centrosymmetric coupling coils, such as rectangular coils. In this paper, a data-driven strategy based on improved ResNet is proposed to recognize the above multi-type misalignments simultaneously, which is accurate, reliable, and of practical and general value. Large-scale data is sampled using the proposed detection coils by double-group cooperation, which features high sensitivity to the misalignments of all kinds, especially yaw angle. Benefit to the compression technique in the pre-processing phase, the number of training data and labels can be successfully reduced by 3/4. As the core algorithm of the overall strategy, the adaptive channel parameter recalibration ResNet (ACPR-ResNet) can effectively perceive slight differences among similar input samples and map them to proper labels. This is achieved by an adaptive operation of nonlinear transformation and recalibrated importance of channel features during the training process. The effectiveness of the proposed strategy is verified experimentally using a 6.6-kW prototype. Within the 24×24 cm range, 95.2% of the test cases have an error of less than 1.7 cm when recognizing horizontal misalignment, and 91.7% of the test cases have an error of less than 1.5° when recognizing yaw angle.

K* Distribution: Evaluating the Latent Space of Deep Neural Networks Using Local Neighborhood Analysis.

Most examinations of neural networks' learned latent spaces typically employ dimensionality reduction techniques such as t-distributed stochastic neighbor embedding (t-SNE) or uniform manifold approximation and projection (UMAP). These methods distort the local neighborhood in the visualization, making it hard to distinguish the structure of a subset of samples in the latent space. In response to this challenge, we introduce the k* distribution and its corresponding visualization technique. This method uses local neighborhood analysis to guarantee the preservation of the structure of sample distributions for individual classes within the subset of the learned latent space. This facilitates easy comparison of different k* distributions, enabling analysis of how various classes are processed by the same neural network. Our study reveals three distinct distributions of samples within the learned latent space subset: 1) fractured; 2) overlapped; and 3) clustered, providing a more profound understanding of the existing contemporary visualizations. Experiments show that the distribution of samples within the network's learned latent space significantly varies depending on the class. Furthermore, we illustrate that our analysis can be applied to explore the latent space of diverse neural network architectures, various layers within neural networks, transformations applied to input samples, and the distribution of training and testing data for neural networks. Thus, the k* distribution should aid in visualizing the structure inside neural networks and further foster their understanding.

Deep Adversarial Inconsistent Cognitive Sampling for Multiview Progressive Subspace Clustering.

Deep multiview clustering methods have achieved remarkable performance. However, all of them failed to consider the difficulty labels (uncertainty of ground truth for training samples) over multiview samples, which may result in a nonideal clustering network for getting stuck into poor local optima during the training process; worse still, the difficulty labels from the multiview samples are always inconsistent, and such a fact makes it even more challenging to handle. In this article, we propose a novel deep adversarial inconsistent cognitive sampling (DAICS) method for multiview progressive subspace clustering. A multiview binary classification (easy or difficult) loss and a feature similarity loss are proposed to jointly learn a binary classifier and a deep consistent feature embedding network, throughout an adversarial minimax game over difficulty labels of multiview consistent samples. We develop a multiview cognitive sampling strategy to select the input samples from easy to difficult for multiview clustering network training. However, the distributions of easy and difficult samples are mixed together, hence not trivial to achieve the goal. To resolve it, we define a sampling probability with a theoretical guarantee. Based on that, a golden section mechanism is further designed to generate a sample set boundary to progressively select the samples with varied difficulty labels via a gate unit, which is utilized to jointly learn a multiview common progressive subspace and clustering network for more efficient clustering. Experimental results on four real-world datasets demonstrate the superiority of DAICS over state-of-the-art methods.

An ECG Deep Learning user identification architecture using ECG sex recognition as a selective parameter

Background:Human user authentication can be implemented by token-, keyword-, or identity-based mechanisms for digital environment session entry (i.e., smartphones, platforms with log-in). Physiological signals, such as ECG, have shown discriminative properties for user identity recognition. Due to ECG hidden nature, it is resilient to public trait exposition, light/noise saturation, or eavesdropping in contrast to fingerprint, facial, voice, or password approaches. ECG might fill those gaps toward a cooperative authentication environment. Methods:This paper proposes a Deep Learning identification scenario in which the inclusion of sex recognition directs the input sample toward a sex-specialized identity classification model, simplifying the discrimination space for each model. The architecture proposed could be suitable for large populations. Our scheme worked with an ECG three-axis pseudo-orthogonal configuration in which each axis is transformed into a time-frequency space. Additionally, we combine each lead matrix in an RGB image, joining the contribution of each wavelet waveform. Results:Our results suggest that it is possible to identify people by using RGB wavelet representations, achieving a classification average of 99.97%. In addition, the inclusion of the sex category for our identification purpose does not significantly affect the classification performance, making it a feasible solution for systems with a larger population. Conclusions:With the features of our database, we have evidence that it is possible to recognize a person’s identity using an ECG sex recognition module through our proposed architecture.

Single Cell Micro RNA Sequencing Library Preparation.

Micro RNAs represent important post-transcriptional regulators in health and are involved in the onset of many diseases. Therefore, the further characterization of physiological miRNA functions is an important basic research question, and miRNAs even have high potential as biomarkers both for prognosis and diagnosis. In order to exploit this potential, it is mandatory to accurately quantify the miRNA expression not only in bulk but also on the single-cell level. Here, we describe a protocol, which facilitates miRNA sequencing library preparation of very low input samples, single cells, and even clinical samples such as circulating tumor cells. The protocol can be combined with different single-cell isolation methods (e.g., micromanipulation and FACS sorting). After cell lysis, sequencing adapters are ligated to the miRNAs, other ncRNA species, and adapter dimers are reduced by exonuclease digest, the miRNA library is reverse transcribed, amplified, and purified. Furthermore, quality controls are described to select only high-quality samples for sequencing.

Image Captioning via Dynamic Path Customization.

This article explores a novel dynamic network for vision and language (V&L) tasks, where the inferring structure is customized on the fly for different inputs. Most previous state-of-the-art (SOTA) approaches are static and handcrafted networks, which not only heavily rely on expert knowledge but also ignore the semantic diversity of input samples, therefore resulting in suboptimal performance. To address these issues, we propose a novel Dynamic Transformer Network (DTNet) for image captioning, which dynamically assigns customized paths to different samples, leading to discriminative yet accurate captions. Specifically, to build a rich routing space and improve routing efficiency, we introduce five types of basic cells and group them into two separate routing spaces according to their operating domains, i.e., spatial and channel. Then, we design a Spatial-Channel Joint Router (SCJR), which endows the model with the capability of path customization based on both spatial and channel information of the input sample. To validate the effectiveness of our proposed DTNet, we conduct extensive experiments on the MS-COCO dataset and achieve new SOTA performance on both the Karpathy split and the online test server. The source code is publicly available at https://github.com/xmu-xiaoma666/DTNet.

An improved neural network for modeling airfoil's unsteady aerodynamics in transonic flow

Understanding the aerodynamic hysteresis loop phenomenon is essential when assessing aerodynamic performance and designing aircraft control systems. This phenomenon is a result of time delay effects in unsteady flow. Traditional methods of predicting unsteady aerodynamic forces using computational fluid dynamics have drawbacks, such as long cycles and low efficiency. In this paper, we focus on predicting the aerodynamic hysteresis loop of the NACA (National Advisory Committee for Aeronautics) 0012 airfoil in transonic flow using a new model called LIDNN (Latin hypercube sample input deep neural network). This model integrates input signals and optimization methods to improve upon traditional neural network models. Based on the example validation, the LIDNN model is authenticated as an accurate and efficient method in predicting the unsteady aerodynamic hysteresis loop of the NACA 0012 airfoil in transonic flow, and another significant advantage of the proposed model is its ability to solve multivariable problems effectively, even under varying Mach numbers.

Design and analysis of thermoelectric energy harvester with 85.6% peak end-to-end efficiency and 38 mV self-startup using 4.13 mΩ/°C indirect temperature dependent MPPT

We present an efficient power management system (PMS) to harvest energy from thermoelectric energy generators (TEGs) via indirect temperature tracking. The PMS includes a power converter with an indirect temperature-dependent (ITD) maximum power point tracking (MPPT) controller and two complementary self-startups. The ITD-MPPT controller consists of an ITD input sampling circuit (ITD-ISC) and an ITD impedance matching circuit (ITD-IMC). The ITD-ISC detects and samples the TEG voltage (VTEG) under dynamically varying temperature gradients. The ITD-ISC removes unnecessary detachments of the converter from multi-mode reconfigurable TEGs to recover the otherwise wasted power using an adaptive input sampling (AIS) technique. The ITD-IMC adjusts the on-time of the power switch to control the input impedance (RIN) of the converter with a relatively high resolution of 4.13 mΩ/°C. This approach provides an ultrawide range source tracking from 2 Ω to 200 Ω. We also present a precision zero current switching (ZCS) controller to manage the on-time of the two power switches simultaneously. Two complementary self-startup components are realized using transformer-based (TSC) and charge-pump-based startup circuits (CSC). Integrated circuits for the PMS are fabricated using 180 nm and 28 nm CMOS processes for high efficiency and low voltage, respectively. The measurements indicate that the proposed PMS achieves an end-to-end efficiency (ηEE) above 80% over a wide source VTEG range (100 mV – 450 mV). The peak ηEE of 85.6% is achieved for VTEG = 200 mV, demonstrating the high-performance MPPT using indirect temperature tracking. TSC and CSC demonstrate successful self-startups at input voltages of 38 mV and 200 mV, respectively, thus providing reliable startup methods for various renewable energy sources.