Weeds detection efficiency through different convolutional neural networks technology

The preservation of the environment has become a priority and a subject that is receiving more and more attention. This is particularly important in the field of precision agriculture, where pesticide and herbicide use has become more controlled. In this study, we propose to evaluate the ability of the deep learning (DL) and convolutional neural network (CNNs) technology to detect weeds in several types of crops using a perspective and proximity images to enable localized and ultra-localized herbicide spraying in the region of Beni Mellal in Morocco. We studied the detection of weeds through six recent CNN known for their speed and precision, namely, VGGNet (16 and 19), GoogLeNet (Inception V3 and V4) and MobileNet (V1 and V2). The first experiment was performed with the CNNs architectures from scratch and the second experiment with their pre-trained versions. The results showed that Inception V4 achieved the highest precision with a rate of 99.41% and 99.51% on the mixed image sets and for its version from scratch and its pre-trained version respectively, and that MobileNet V2 was the fastest and lightest with its size of 14 MB.

A Novel Machine Learning Approach for Sentiment Analysis on Twitter Incorporating the Universal Language Model Fine-Tuning and SVM

Twitter sentiment detectors (TSDs) provide a better solution to evaluate the quality of service and product than other traditional technologies. The classification accuracy and detection performance of TSDs, which are extremely reliant on the performance of the classification techniques, are used, and the quality of input features is provided. However, the time required is a big problem for the existing machine learning methods, which leads to a challenge for all enterprises that aim to transform their businesses to be processed by automated workflows. Deep learning techniques have been utilized in several real-world applications in different fields such as sentiment analysis. Deep learning approaches use different algorithms to obtain information from raw data such as texts or tweets and represent them in certain types of models. These models are used to infer information about new datasets that have not been modeled yet. We present a new effective method of sentiment analysis using deep learning architectures by combining the “universal language model fine-tuning” (ULMFiT) with support vector machine (SVM) to increase the detection efficiency and accuracy. The method introduces a new deep learning approach for Twitter sentiment analysis to detect the attitudes of people toward certain products based on their comments. The extensive results on three datasets illustrate that our model achieves the state-of-the-art results over all datasets. For example, the accuracy performance is 99.78% when it is applied on the Twitter US Airlines dataset.

STEEL enables high-resolution delineation of spatiotemporal transcriptomic data

Advances in spatial transcriptomics enlarge the use of single cell technologies to unveil the expression landscape of the tissues with valuable spatial context. However, computational tools developed for single-cell transcriptomics have great limits in dealing with spatial transcriptomic data with high noise on detected transcript signals. Here we propose an unsupervised and manifold learning-based algorithm, STEEL, which identifies different cell types from spatial transcriptome by clustering cells/beads exhibiting both highly similar gene expression profiles and close spatial distance in the manner of graphs. Comprehensive evaluation of STEEL on various spatial transcriptomic datasets from 10X Visium platform demonstrates that it not only achieves a high resolution to characterize fine structures of mouse brain, but also enables the integration of multiple tissue slides individually analyzed into a larger one. STEEL outperforms previous methods to effectively distinguish different cell types of various tissues on Slide-seq datasets, featuring in higher bead density but lower transcript detection efficiency. Application of STEEL on spatial transcriptomes of early-stage mouse embryos (E9.5 to E12.5) successfully delineates a progressive development landscape of tissues from ectoderm, mesoderm and endoderm layers, and futher profiles dynamic changes on cell differentiation in heart and other organs. With the advancement of spatial transcriptome technologies, our method will have great applicability in high-resolution cell type identification and unbiased spatiotemporal data integration.

Development of 222Rn Emanation Sources with Integrated Quasi 2π Active Monitoring

In this work, a novel approach for the standardization of low-level 222Rn emanation is presented. The technique is based on the integration of a 222Rn source, directly, with an α-particle detector, which allows the residual 222Rn to be continuously monitored. Preparation of the device entails thermal physical vapor deposition of 226RaCl2 directly onto the surface of a commercially available ion implanted Si-diode detector, resulting in a thin-layer geometry. This enables continuous collection of well resolved α-particle spectra of the nuclei, decaying within the deposited layer, with a detection efficiency of approximately 0.5 in a quasi 2π geometry. The continuously sampled α-particle spectra are used to derive the emanation by statistical inversion. It is possible to achieve this with high temporal resolution due to the small background and the high counting efficiency of the presented technique. The emanation derived in this way exhibits a dependence on the relative humidity of up to 15% in the range from 20% rH to 90% rH. Traceability to the SI is provided by employing defined solid-angle α-particle spectrometry to characterize the counting efficiency of the modified detectors. The presented technique is demonstrated to apply to a range covering the release of at least 1 to 210 222Rn atoms per second, and it results in SI-traceable emanation values with a combined standard uncertainty not exceeding 2%. This provides a pathway for the realization of reference atmospheres covering typical environmental 222Rn levels and thus drastically improves the realization and the dissemination of the derived unit of the activity concentration concerning 222Rn in air.

Data Mining Algorithm for Moving Object Trajectory

Aiming at the poor effect and long recognition time of data mining algorithm for moving target trajectory recognition, a data mining algorithm based on improved Hausdorff distance is proposed. The position and angle of abnormal trajectory data are detected by calculating the distance between trajectory classification and sub trajectory line segments, and the trajectory unit is established by using the improved Hausdorff distance algorithm to optimize the similarity matching structure. Experimental results show that the algorithm has low error pruning rate in identifying moving target trajectory, improves the detection efficiency of moving target trajectory recognition data, and ensures the quality of moving target trajectory recognition data mining

Study of 222−220Rn Measurement Systems Based on Electrostatic Collection by Using Geant4+COMSOL Simulation

Using Monte Carlo (with Geant4) and COMSOL simulations, the authors have defined a useful tool to reproduce the alpha spectroscopy of 222Rn, 220Rn and their ionized daughters by measurement systems based on electrostatic collection on a silicon detector, inside a metallic chamber. Several applications have been performed: (i) simulating commercial devices worldwide used, and comparing them with experimental theoretical results; (ii) studying of realization of new measurement systems through investigation of the detection efficiency versus different chamber geometries. New considerations and steps forward have been drawn. The present work is a novelty in the literature concerning this research framework.

The Rapid Antigen Detection Test for SARS-CoV-2 Underestimates the Identification of COVID-19 Positive Cases and Compromises the Diagnosis of the SARS-CoV-2 (K417N/T, E484K, and N501Y) Variants

Timely detection of severe acute respiratory syndrome due to coronavirus 2 (SARS-CoV-2) by reverse transcription quantitative polymerase chain reaction (RT-qPCR) has been the gold- strategy for identifying positive cases during the current pandemic. However, faster and less expensive methodologies are also applied for the massive diagnosis of COVID-19. In this way, the rapid antigen test (RAT) is widely used. However, it is necessary to evaluate its detection efficiency considering the current pandemic context with the circulation of new viral variants. In this study, we evaluated the sensitivity and specificity of RAT (SD BIOSENSOR, South Korea), widely used for testing and SARS-CoV-2 diagnosis in Santiago of Chile. The RAT showed a 90% (amplification range of 20 ≤ Cq <25) and 10% (amplification range of 25 ≤ Cq <30) of positive SARS-CoV-2 cases identified previously by RT-qPCR. Importantly, a 0% detection was obtained for samples within a Cq value>30. In SARS-CoV-2 variant detection, RAT had a 42.8% detection sensitivity in samples with RT-qPCR amplification range 20 ≤ Cq <25 containing the single nucleotide polymorphisms (SNP) K417N/T, N501Y and E484K, associated with beta or gamma SARS-CoV-2 variants. This study alerts for the special attention that must be paid for the use of RAT at a massive diagnosis level, especially in the current scenario of appearance of several new SARS-CoV-2 variants which could generate false negatives and the compromise of possible viral outbreaks.

Investigation of parameters of new MAPD-3NM silicon photomultipliers

In the presented work, the parameters of a new MAPD-3NM-II photodiode with buried pixel structure manufactured in cooperation with Zecotek Company are investigated. The photon detection efficiency, gain, capacitance and gamma-ray detection performance of photodiodes are studied. The SPECTRIG MAPD is used to measure the parameters of the MAPD-3NM-II and scintillation detector based on it. The obtained results show that the newly developed MAPD-3NM-II photodiode outperforms its counterparts in most parameters and it can be successfully applied in space application, medicine, high-energy physics and security.

Automated SEM Image Analysis of the Sphere Diameter, Sphere-Sphere Separation, and Opening Size Distributions of Nanosphere Lithography Masks

Colloidal nanosphere monolayers—used as a lithography mask for site-controlled material deposition or removal—offer the possibility of cost-effective patterning of large surface areas. In the present study, an automated analysis of scanning electron microscopy (SEM) images is described, which enables the recognition of the individual nanospheres in densely packed monolayers in order to perform a statistical quantification of the sphere size, mask opening size, and sphere-sphere separation distributions. Search algorithms based on Fourier transformation, cross-correlation, multiple-angle intensity profiling, and sphere edge point detection techniques allow for a sphere detection efficiency of at least 99.8%, even in the case of considerable sphere size variations. While the sphere positions and diameters are determined by fitting circles to the spheres edge points, the openings between sphere triples are detected by intensity thresholding. For the analyzed polystyrene sphere monolayers with sphere sizes between 220 and 600 nm and a diameter spread of around 3% coefficients of variation of 6.8–8.1% for the opening size are found. By correlating the mentioned size distributions, it is shown that, in this case, the dominant contribution to the opening size variation stems from nanometer-scale positional variations of the spheres.