Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Pattern Recognition”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 235 records · Page 13

NMR-based metabonomic analysis of the hepatotoxicity induced by combined exposure to PCBs and TCDD in rats

A metabonomic approach using {sup 1}H NMR spectroscopy was adopted to investigate the metabonomic pattern of rat urine after oral administration of environmental endocrine disruptors (EDs) polychlorinated biphenyls (PCBs) and 2,3,7,8-tetrachlorodibenzo- p-dioxin (TCDD) alone or in combination and to explore the possible hepatotoxic mechanisms of combined exposure to PCBs and TCDD. {sup 1}H NMR spectra of urines collected 24 h before and after exposure were analyzed via pattern recognition by using principal component analysis (PCA). Serum biochemistry and liver histopathology indicated significant hepatotoxicity in the rats of the combined group. The PCA scores plots of urinary {sup 1}H NMR data showed that all the treatment groups could be easily distinguished from the control group, so could the PCBs or TCDD group and the combined group. The loadings plots of the PCA revealed remarkable increases in the levels of lactate, glucose, taurine, creatine, and 2-hydroxy-isovaleric acid and reductions in the levels of 2-oxoglutarate, citrate, succinate, hippurate, and trimethylamine-N-oxide in rat urine after exposure. These changes were more striking in the combined group. The changed metabolites may be considered possible biomarker for the hepatotoxicity. The present study demonstrates that combined exposure to PCBs and TCDD induced significant hepatotoxicity in rats, and mitochondrial dysfunction and fatty acid metabolism perturbations might contribute to the hepatotoxicity. There was good conformity between changes in the urine metabonomic pattern and those in serum biochemistry and liver histopathology. These results showed that the NMR-based metabonomic approach may provide a promising technique for the evaluation of the combined toxicity of EDs.

60 APPLIED LIFE SCIENCES↗

Hepatic lipid profiling of deer mice fed ethanol using {sup 1}H and {sup 31}P NMR spectroscopy: A dose-dependent subchronic study

Chronic alcohol abuse is a 2nd major cause of liver disease resulting in significant morbidity and mortality. Alcoholic liver disease (ALD) is characterized by a wide spectrum of pathologies starting from fat accumulation (steatosis) in early reversible stage to inflammation with or without fibrosis and cirrhosis in later irreversible stages. Previously, we reported significant steatosis in the livers of hepatic alcohol dehydrogenase (ADH)-deficient (ADH{sup −}) vs. hepatic ADH-normal (ADH{sup +}) deer mice fed 4% ethanol daily for 2 months [Bhopale et al., 2006, Alcohol 39, 179–188]. However, ADH{sup −} deer mice fed 4% ethanol also showed a significant mortality. Therefore, a dose-dependent study was conducted to understand the mechanism and identify lipid(s) involved in the development of ethanol-induced fatty liver. ADH{sup −} and ADH{sup +} deer mice fed 1, 2 or 3.5% ethanol daily for 2 months and fatty infiltration in the livers were evaluated by histology and by measuring dry weights of extracted lipids. Lipid metabolomic changes in extracted lipids were determined by proton ({sup 1}H) and {sup 31}phosphorus ({sup 31}P) nuclear magnetic resonance (NMR) spectroscopy. The NMR data was analyzed by hierarchical clustering (HC) and principle component analysis (PCA) for pattern recognition. Extensive vacuolization by histology and significantly increased dry weights of total lipids found only in the livers of ADH{sup −} deer mice fed 3.5% ethanol vs. pair-fed controls suggest a dose-dependent formation of fatty liver in ADH{sup −} deer mouse model. Analysis of NMR data of ADH{sup −} deer mice fed 3.5% ethanol vs. pair-fed controls shows increases for total cholesterol, esterified cholesterol, fatty acid methyl esters (FAMEs), triacylglycerides and unsaturation, and decreases for free cholesterol, phospholipids and allylic and diallylic protons. Certain classes of neutral lipids (cholesterol esters, fatty acyl chain (-COCH{sub 2}-) and FAMEs) were also mildly increased in ADH{sup −} deer mice fed 1 or 2% ethanol. Only small increases were observed for allylic and diallylic protons, FAMEs and unsaturations in ADH{sup +} deer mice fed 3.5% ethanol vs. pair-fed controls. PCA of NMR data showed increased clustering by gradual separation of ethanol-fed ADH{sup −} deer mice groups from their respective pair-fed control groups and corresponding ethanol-fed ADH{sup +} deer mice groups. Our data indicate that dose of ethanol and hepatic ADH deficiency are two key factors involved in initiation and progression of alcoholic fatty liver disease. Further studies on characterization of individual lipid entities and associated metabolic pathways altered in our deer mouse model after different durations of ethanol feeding could be important to delineate mechanism(s) and identify potential biomarker candidate(s) of early stage ALD. -- Highlights: ► Dose-dependent ethanol-induced fatty liver was studied in deer mouse model. ► A NMR-based lipidomic approach with histology and dry lipid weights was used. ► We used principal component analysis (PCA) to analyze the NMR lipidomic data. ► Dose-dependent clustering patterns by PCA were compared among the groups.

60 APPLIED LIFE SCIENCES↗

ClimateNet: an expert-labeled open dataset and deep learning architecture for enabling high-precision analyses of extreme weather

Abstract. Identifying, detecting, and localizing extreme weather events is a crucial first step in understanding how they may vary under different climate change scenarios. Pattern recognition tasks such as classification, object detection, and segmentation (i.e., pixel-level classification) have remained challenging problems in the weather and climate sciences. While there exist many empirical heuristics for detecting extreme events, the disparities between the output of these different methods even for a single event are large and often difficult to reconcile. Given the success of deep learning (DL) in tackling similar problems in computer vision, we advocate a DL-based approach. DL, however, works best in the context of supervised learning – when labeled datasets are readily available. Reliable labeled training data for extreme weather and climate events is scarce. We create “ClimateNet” – an open, community-sourced human-expert-labeled curated dataset that captures tropical cyclones (TCs) and atmospheric rivers (ARs) in high-resolution climate model output from a simulation of a recent historical period. We use the curated ClimateNet dataset to train a state-of-the-art DL model for pixel-level identification – i.e., segmentation – of TCs and ARs. We then apply the trained DL model to historical and climate change scenarios simulated by the Community Atmospheric Model (CAM5.1) and show that the DL model accurately segments the data into TCs, ARs, or “the background” at a pixel level. Further, we show how the segmentation results can be used to conduct spatially and temporally precise analytics by quantifying distributions of extreme precipitation conditioned on event types (TC or AR) at regional scales. The key contribution of this work is that it paves the way for DL-based automated, high-fidelity, and highly precise analytics of climate data using a curated expert-labeled dataset – ClimateNet. ClimateNet and the DL-based segmentation method provide several unique capabilities: (i) they can be used to calculate a variety of TC and AR statistics at a fine-grained level; (ii) they can be applied to different climate scenarios and different datasets without tuning as they do not rely on threshold conditions; and (iii) the proposed DL method is suitable for rapidly analyzing large amounts of climate model output. While our study has been conducted for two important extreme weather patterns (TCs and ARs) in simulation datasets, we believe that this methodology can be applied to a much broader class of patterns and applied to observational and reanalysis data products via transfer learning.

54 ENVIRONMENTAL SCIENCES↗

A decision tree algorithm for investigation of model biases related to dynamical cores and physical parameterizations: CESM/CAM EVALUATION BY DECISION TREES

An object-based evaluation method using a pattern recognition algorithm (i.e., classification trees) is applied to the simulated orographic precipitation for idealized experimental setups using the National Center of Atmospheric Research (NCAR) Community Atmosphere Model (CAM) with the finite volume (FV) and the Eulerian spectral transform dynamical cores with varying resolutions. Daily simulations were analyzed and three different types of precipitation features were identified by the classification tree algorithm. The statistical characteristics of these features (i.e., maximum value, mean value, and variance) were calculated to quantify the difference between the dynamical cores and changing resolutions. Even with the simple and smooth topography in the idealized setups, complexity in the precipitation fields simulated by the models develops quickly. The classification tree algorithm using objective thresholding successfully detected different types of precipitation features even as the complexity of the precipitation field increased. The results show that the complexity and the bias introduced in small-scale phenomena due to the spectral transform method of CAM Eulerian spectral dynamical core is prominent, and is an important reason for its dissimilarity from the FV dynamical core. The resolvable scales, both in horizontal and vertical dimensions, have significant effect on the simulation of precipitation. The results of this study also suggest that an efficient and informative study about the biases produced by GCMs should involve daily (or even hourly) output (rather than monthly mean) analysis over local scales.

54 ENVIRONMENTAL SCIENCES↗

Determining Full Three-Dimensional Orientation of Single Anisotropic Nanoparticles by Differential Interference Contrast Microscopy

Keeping track: By combining differential interference contrast (DIC) image pattern recognition with DIC polarization anisotropy, the exact full three-dimensional angular information of individual tilted gold nanorods positioned in the focal plane of the objective lens can be readily determined. The angular rotational modes and kinetics of individual in-focus gold nanorods can thus be resolved dynamically.

36 MATERIALS SCIENCE↗

Electroluminescence TPCs at the thermal diffusion limit

Here, the NEXT experiment aims at searching for the hypothetical neutrinoless double-beta decay from the ${}^{136}$Xe isotope using a high-purity xenon TPC. Efficient discrimination of the events through pattern recognition of the topology of primary ionisation tracks is a major requirement for the experiment. However, it is limited by the diffusion of electrons. It is known that the addition of a small fraction of a molecular gas to xenon reduces electron diffusion. On the other hand, the electroluminescence (EL) yield drops and the achievable energy resolution may be compromised. We have studied the effect of adding several molecular gases to xenon (CO${}_{2}$, CH${}_{4}$ and CF${}_{4}$) on the EL yield and energy resolution obtained in a small prototype of driftless gas proportional scintillation counter. We have compared our results on the scintillation characteristics (EL yield and energy resolution) with a microscopic simulation, obtaining the diffusion coefficients in those conditions as well. Accordingly, electron diffusion may be reduced from about 10 mm/$\sqrt{\mathrm{m}}$ for pure xenon down to 2.5 mm/$\sqrt{\mathrm{m}}$ using additive concentrations of about 0.05%, 0.2% and 0.02% for CO${}_{2}$, CH${}_{4}$ and CF${}_{4}$, respectively. Our results show that CF${}_{4}$ admixtures present the highest EL yield in those conditions, but very poor energy resolution as a result of huge fluctuations observed in the EL formation. CH${}_{4}$ presents the best energy resolution despite the EL yield being the lowest. The results obtained with xenon admixtures are extrapolated to the operational conditions of the NEXT-100 TPC. CO${}_{2}$ and CH${}_{4}$ show potential as molecular additives in a large xenon TPC, CH${}_{4}$ showing the best performance and stability to be used in the NEXT-100 TPC, with an extrapolated energy resolution of 0.4% at 2.45 MeV for concentrations below 0.4%.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Eigenparticles: characterizing particles using eigenfaces

The shape characteristics of particles have a pinnacle role in microsopic and macroscopic features of a system. Several studies have highlighted the need for considering deviations from a spherical representation of particles for accurate modeling of granular and multiphase flow systems. Using a shape factor, sphericity or roundness parameter alone is proven to be inadequate to capture the physical phenomena. In the present study we propose a novel metric based on the pattern recognition method Eigenfaces, coining the technique ‘Eigenparticles’. Using this technique we create a single statistical distribution of basis shapes to describe the morphological composition. The proposed technique is successfully validated with test shapes and applied to real particles. As a result, when compared with a state-of-the-art Fourier based method, ‘Eigenparticles’ performs favorably, clearly distinguishing the different particles.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Distributed non-negative matrix factorization with determination of the number of latent features

The holistic analysis and understanding of the latent (that is, not directly observable) variables and patterns buried in large datasets is crucial for data-driven science, decision making and emergency response. Such exploratory analyses require devising unsupervised learning methods for data mining and extraction of the latent features, and non-negative matrix factorization (NMF) is one of the prominent such methods. NMF is based on compute-intense non-convex constrained minimization, which, for large datasets requires fast and distributed algorithms. However, current parallel implementations of NMF fail to estimate the number of latent features. In practice, identifying these features is both difficult and significant for pattern recognition and latent feature analysis, especially for large dense matrices. Here, we introduce a distributed NMF algorithm coupled with distributed custom clustering followed by a stability analysis on dense data, which we call DnMFk, to determine the number of latent variables. The results on synthetic data and the classical Swimmer data set demonstrate the accuracy of model determination while scaling nearly linearly across multiple processors for large data. Further, we employ DnMFk to determine the number of hidden features from a terabyte matrix.

97 MATHEMATICS AND COMPUTING↗

Multiscale Persistent Functions for Biomolecular Structure Characterization

Here in this paper, we introduce multiscale persistent functions for biomolecular structure characterization. The essential idea is to combine our multiscale rigidity functions (MRFs) with persistent homology analysis, so as to construct a series of multiscale persistent functions, particularly multiscale persistent entropies, for structure characterization. To clarify the fundamental idea of our method, the multiscale persistent entropy (MPE) model is discussed in great detail. Mathematically, unlike the previous persistent entropy (Chintakunta et al. in Pattern Recognit 48(2):391–401, 2015; Merelli et al. in Entropy 17(10):6872–6892, 2015; Rucco et al. in: Proceedings of ECCS 2014, Springer, pp 117–128, 2016), a special resolution parameter is incorporated into our model. Various scales can be achieved by tuning its value. Physically, our MPE can be used in conformational entropy evaluation. More specifically, it is found that our method incorporates in it a natural classification scheme. This is achieved through a density filtration of an MRF built from angular distributions. To further validate our model, a systematical comparison with the traditional entropy evaluation model is done. Additionally, it is found that our model is able to preserve the intrinsic topological features of biomolecular data much better than traditional approaches, particularly for resolutions in the intermediate range. Moreover, by comparing with traditional entropies from various grid sizes, bond angle-based methods and a persistent homology-based support vector machine method (Cang et al. in Mol Based Math Biol 3:140–162, 2015), we find that our MPE method gives the best results in terms of average true positive rate in a classic protein structure classification test. More interestingly, all-alpha and all-beta protein classes can be clearly separated from each other with zero error only in our model. Finally, a special protein structure index (PSI) is proposed, for the first time, to describe the “regularity” of protein structures. Basically, a protein structure is deemed as regular if it has a consistent and orderly configuration. Our PSI model is tested on a database of 110 proteins; we find that structures with larger portions of loops and intrinsically disorder regions are always associated with larger PSI, meaning an irregular configuration, while proteins with larger portions of secondary structures, i.e., alpha-helix or beta-sheet, have smaller PSI. Essentially, PSI can be used to describe the “regularity” information in any systems.

59 BASIC BIOLOGICAL SCIENCES↗

IDEAL: Images Across Domains, Experiments, Algorithms and Learning

Research across science domains is increasingly reliant on image-centric data. Software tools are in high demand to uncover relevant, but hidden, information in digital images, such as those coming from faster next generation high-throughput imaging platforms. The challenge is to analyze the data torrent generated by the advanced instruments efficiently, and provide insights such as measurements for decision-making. In this article, we overview work performed by an interdisciplinary team of computational and materials scientists, aimed at designing software applications and coordinating research efforts connecting (1) emerging algorithms for dealing with large and complex datasets; (2) data analysis methods with emphasis in pattern recognition and machine learning; and (3) advances in evolving computer architectures. Engineering tools around these efforts accelerate the analyses of image-based recordings, improve reusability and reproducibility, scale scientific procedures by reducing time between experiments, increase efficiency, and open opportunities for more users of the imaging facilities. This paper describes our algorithms and software tools, showing results across image scales, demonstrating how our framework plays a role in improving image understanding for quality control of existent materials and discovery of new compounds.

97 MATHEMATICS AND COMPUTING↗

A Siamese CNN + KNN-Based Classification Framework for Non-intrusive Load Monitoring

Through the development of smart grids, programs such as demand side response, have been presented as auxiliary services to the real-time operation of distributed networks. In order to provide consumers information on their energy consumption, so that a modulation in consumption is possible, non-intrusive load monitoring has been introduced as an solution to this pattern recognition problem. Non-intrusive load monitoring enables the modeling of electrical loads connected to the low-voltage system, considering only a single measurement point. Presented state-of-the-art solutions though, consider availability of data as well as representation of all possible classes of the environment. This is of course a most conservative hypothesis, since in real-life applications availability of such data is much difficult, as well as the dynamic behavior of models is implicitly evolving in time. Here, a framework that uses neural Siamese networks with k-nearest neighbor clustering is presented toward non-intrusive load monitoring. Online learning feature is implemented, which relaxes the hypothesis of data requirements as well addresses the evolving nature of load profile. k-nearest clustering allows nonlinear characteristic space modelling. Test results using synthetics and real-life data show that the solution, besides obtaining a good generalizability in the classification, also obtained results with an accuracy of 95.77%.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Quantum computing based hybrid deep learning for fault diagnosis in electrical power systems

Quantum computing (QC) and deep learning have shown promise of supporting transformative advances and have recently gained popularity in a wide range of areas. Here, this paper proposes a hybrid QC-based deep learning framework for fault diagnosis of electrical power systems that combine the feature extraction capabilities of conditional restricted Boltzmann machine with an efficient classification of deep networks. Computational challenges stemming from the complexities of such deep learning models are overcome by QC-based training methodologies that effectively leverage the complementary strengths of quantum assisted learning and classical training techniques. The proposed hybrid QC-based deep learning framework is tested on a simulated electrical power system with 30 buses and wide variations of substation and transmission line faults, to demonstrate the framework’s applicability, efficiency, and generalization capabilities. High computational efficiency is enjoyed by the proposed hybrid approach in terms of computational effort required and quality of diagnosis performance over classical training methods. In addition, superior and reliable fault diagnosis performance with faster response time is achieved over state-of-the-art pattern recognition methods based on artificial neural networks (ANN) and decision trees (DT).

42 ENGINEERING↗

Endothelial RIG-I activation impairs endothelial function

Highlights: Black-Right-Pointing-Pointer RIG-I activation impairs endothelial function in vivo. Black-Right-Pointing-Pointer RIG-I activation alters HCAEC biology in vitro. Black-Right-Pointing-Pointer EPC function is affected by RIG-I stimulation in vitro. -- Abstract: Background: Endothelial dysfunction is a crucial part of the chronic inflammatory atherosclerotic process and is mediated by innate and acquired immune mechanisms. Recent studies suggest that pattern recognition receptors (PRR) specialized in immunorecognition of nucleic acids may play an important role in endothelial biology in a proatherogenic manner. Here, we analyzed the impact of endothelial retinoic acid inducible gene I (RIG-I) activation upon vascular endothelial biology. Methods and results: Wild type mice were injected intravenously with 32.5 {mu}g of the RIG-ligand 3pRNA (RNA with triphosphate at the 5 Prime end) or polyA control every other day for 7 days. In 3pRNA-treated mice, endothelium-depended vasodilation was significantly impaired, vascular oxidative stress significantly increased and circulating endothelial microparticle (EMP) numbers significantly elevated compared to controls. To gain further insight in RIG-I dependent endothelial biology, cultured human coronary endothelial cells (HCAEC) and endothelial progenitor cells (EPC) were stimulated in vitro with 3pRNA. Both cells types express RIG-I and react with receptor upregulation upon stimulation. Reactive oxygen species (ROS) formation is enhanced in both cell types, whereas apoptosis and proliferation is not significantly affected in HCAEC. Importantly, HCAEC release significant amounts of proinflammatory cytokines in response to RIG-I stimulation. Conclusion: This study shows that activation of the cytoplasmatic nucleic acid receptor RIG-I leads to endothelial dysfunction. RIG-I induced endothelial damage could therefore be an important pathway in atherogenesis.

60 APPLIED LIFE SCIENCES↗

NMR-based serum metabolomics study reveals a innovative diagnostic model for missed abortion

A missed abortion (MA) is an in-utero death of the embryo or fetus before the 20th week of gestation with retained products of conception. In order to discover novel biomarkers for MA, a {sup 1}H NMR spectroscopy-based metabolomics approach was applied to detect human MA serum metabolic profiles. Serum samples were obtained from patients with MA (n = 15) and healthy controls (n = 9) for study. The NOESYPR1D spectrum combined with multi-variate pattern recognition analysis was used to cluster the groups and establish a disease-specific metabolites phenotype. Principal component analysis (PCA) and orthogonal partial least-squares discriminant analysis (OPLS-DA) models were capable of distinguishing MA patients from healthy subjects. The results revealed that 24 metabolites altered in MA patients compared with the control population. Metabolomic pathway analysis demonstrated that alanine, aspartate and glutamate metabolism, citrate cycle (TCA cycle), taurine and hypotaurine metabolism were significantly altered in MA. The results indicated that serum NMR-based metabolomic profiling method is sensitive and specific enough to distinguish MA and from healthy controls, this method could be developed as a clinically useful diagnostic tool for MA. The finding from the MA serum metabolic profiling shed a new light on further understanding of MA disease mechanisms.

60 APPLIED LIFE SCIENCES↗

Vaccine adjuvant ARNAX promotes mucosal IgA production in influenza HA vaccination

Adjuvant stimulates pattern-recognition receptors (PRRs) expressed by dendritic cells, which causes immune-enhancing of T lymphocytes. Adjuvant also induces innate immune response in whole-body cells via PRRs to evoke cytokinemia. A cytokine-mediated immune response is important for the systemic protection of a host from microbial infections. Using an influenza subcomponent vaccine in a mouse model, we intranasally administered a TLR3-specific adjuvant ARNAX + HA split vaccine to mice. ARNAX efficiently induced mucosal IgA and systemic IgG production by nasal drop. Moreover, ARNAX + HA simultaneously induced CD8 and CD4 T cell activation. We have previously shown that ARNAX does not induce harmful systemic cytokine production. Thus, our findings indicate that the ARNAX + HA vaccine is a harmless prophylactic vaccine for flu that induces HA-specific T cell activation and IgA/IgG production. These results suggested that ARNAX + antigen enhanced the immune response without inducing inflammatory toxicity for vaccination against infectious diseases.

60 APPLIED LIFE SCIENCES↗

Dynamic analysis environment for nuclear forensic analyses

We report a Dynamic Analysis Environment (DAE) software package is introduced to facilitate group inclusion/exclusion method testing, evaluation and comparison for pre-detonation nuclear forensics applications. Employing DAE, the multivariate signatures of a questioned material can be compared to the signatures for different, known groups, enabling the linking of the questioned material to its potential process, location, or fabrication facility. Advantages of using DAE for group inclusion/exclusion include built-in query tools for retrieving data of interest from a database, the recording and documentation of all analysis steps, a clear visualization of the analysis steps intelligible to a non-expert, and the ability to integrate analysis tools developed in different programming languages. Two group inclusion/exclusion methods are implemented in DAE: principal component analysis, a parametric feature extraction method, and k nearest neighbors, a nonparametric pattern recognition method. Spent Fuel Isotopic Composition (SFCOMPO), an open source international database of isotopic compositions for spent nuclear fuels (SNF) from 14 reactors, is used to construct PCA and KNN models for known reactor groups, and 20 simulated SNF samples are utilized in evaluating the performance of these group inclusion/exclusion models. For all 20 simulated samples, PCA in conjunction with the Q statistic correctly excludes a large percentage of reactor groups and correctly includes the true reactor of origination. As a result, employing KNN, 14 of the 20 simulated samples are classified to their true reactor of origination.

97 MATHEMATICS AND COMPUTING↗

Power quality disturbances diagnosis: A 2D densely connected convolutional network framework

The fast and accurate diagnosis of power quality disturbances (PQD) aids in avoiding shutdowns and unnecessary procedures, concerning electric energy distribution systems. As such, a number of techniques have been tested and applied in order to reach this objective. Majority of the techniques applied are two-step based. On the first step, power quality disturbances features are extracted. Second step, considering features extracted, disturbance classification is implemented. Recently, relevant literature has presented data-driven signal processing-based approaches, as deep convolutional neural networks (DCNN), which can implement both processing steps while providing automated recognition of patterns and outliers in data. However, not considered by state-of-art, power quality disturbances are evolving in nature, while all possible regularities might not be represented in the dataset. In this work a 2 Dimension Densely Connected Convolutional Network (2D-DenseNet) framework is presented. Further, a case study with synthetic disturbance events are analyzed. Easy-to-implement formulation, built on the 2D-DenseNet, without hard-to-design parameters, highlight potential aspects for real-life implementation.

42 ENGINEERING↗