Search

vmax, rmw and r17 in experiment CNTL and CNTL_B03.

3D hand pose estimation from single depth is a fundamental problem in computer vision, and has wide applications.However, the existing methods still can not achieve satisfactory hand pose estimation results due to view variation and occlusion of human hand. In this paper, we propose a new virtual view selection and fusion module for 3D hand pose estimation from single depth.We propose to automatically select multiple virtual viewpoints for pose estimation and fuse the results of all and find this empirically delivers accurate and robust pose estimation. In order to select most effective virtual views for pose fusion, we evaluate the virtual views based on the confidence of virtual views using a light-weight network via network distillation. Experiments on three main benchmark datasets including NYU, ICVL and Hands2019 demonstrate that our method outperforms the state-of-the-arts on NYU and ICVL, and achieves very competitive performance on Hands2019-Task1, and our proposed virtual view selection and fusion module is both effective for 3D hand pose estimation., Comment: Accepted by AAAI2022

A novel Fe-Cr-B-C-Al-Si-Mn composite coating was designed and synthesized on 2Cr13 steel substrate using laser cladding technique in this study. The microstructure evolution, microhardness and wear resistance of the composite coating were investigated. The results showed that the microstructure of the coating was mainly composed of α-(Fe,Cr,Al) martensite, retained γ-Fe, and the eutectic comprised of CrFeB, M3(C,B)2, M2(B,C) and M23(C,B)6. The average microhardness of the coating (742.1 HV0.5) was 3.95 times compared with the 2Cr13 steel substrate, and the wear resistance was enhanced by 12.5 times compared with the 2Cr13 steel substrate. The wear mechanisms were dominated by abrasive wear and oxidative wear.

Colorectal cancer (CRC) is the third most common malignancy and the second most common cause of cancer-related mortality worldwide. A total of 20% of CRC patients present with distant metastasis. The hepatic portal venous system, responsible for collecting most intestinal blood, makes the liver the most common site of CRC metastasis. The formation of liver metastases from colorectal cancer is a long and complex process. It involves the maintenance of primary tumors, vasculature invasion, distant colonization, and metastasis formation. In this review, we serve on how the CRC cells acquire stemness, invade the vascular, and colonize the liver. In addition, we highlight how the resident cells of the liver and immune cells interact with CRC cells. We also discuss the current immunotherapy approaches and challenges we face, and finally, we look forward to finding new therapeutic targets based on novel sequencing technologies.

Data for picture in Paper, including low-level wind, RMW, R17, tangential wind budget, SE drived vertical velocity.

Automatic code summarization is beneficial to daily software development since it could help reduce the requirement of manual writing. Currently, artificial intelligence is undergoing a paradigm shift. The foundation models pretrained on massive data and finetuned to downstream tasks surpass specially customized models. This trend inspired us to consider reusing foundation models instead of learning from scratch. Thereby, we propose a flexible and robust approach for automatic code summarization, based on neural models. We assemble available foundation models, such as CodeBERT and GPT-2, into a single neural model named AdaMo. Moreover, we utilize Gaussian noise as the simulation of contextual information to optimize the latent representation. Furthermore, we introduce two adaptive schemes from the perspective of knowledge transfer, namely continuous pretraining and intermediate finetuning, and design intermediate stage tasks for general sequence-to-sequence learning. Finally, we evaluate AdaMo against a benchmark dataset for code summarization, by comparing it with state-of-the-art models.

Automatic code summarization is beneficial to daily software development since it could help reduce the requirement of manual writing. Currently, artificial intelligence is undergoing a paradigm shift. The foundation models pretrained on massive data and finetuned to downstream tasks surpass specially customized models. This trend inspired us to consider reusing foundation models instead of learning from scratch. Thereby, we propose a flexible and robust approach for automatic code summarization, based on neural models. We assemble available foundation models, such as CodeBERT and GPT-2, into a single neural model named AdaMo. Moreover, we utilize Gaussian noise as the simulation of contextual information to optimize the latent representation. Furthermore, we introduce two adaptive schemes from the perspective of knowledge transfer, namely continuous pretraining and intermediate finetuning, and design intermediate stage tasks for general sequence-to-sequence learning. Finally, we evaluate AdaMo against a benchmark dataset for code summarization, by comparing it with state-of-the-art models.

Motivated by the experimental progress in the study of heavy baryons, we investigate the mass spectra of strange single heavy baryons in the $λ$-mode, where the relativistic quark model and the infinitesimally shifted Gaussian basis function method are employed. It is shown that the experimental data can be well reproduced by the predicted masses. The root mean square radii and radial probability density distributions of the wave functions are analyzed in detail. Meanwhile, the mass spectra allow us to successfully construct the Regge trajectories in the $(J,M^{2})$ plane. We also preliminarily assign quantum numbers to the recently observed baryons, including $Ξ_{c}(3055)$, $Ξ_{c}(3080)$, $Ξ_{c}(2930)$, $Ξ_{c}(2923)$, $Ξ_{c}(2939)$, $Ξ_{c}(2965)$, $Ξ_{c}(2970)$, $Ξ_{c}(3123)$, $Ξ_{b}(6100)$, $Ξ_{b}(6227)$, $Ξ_{b}(6327)$ and $Ξ_{b}(6333)$. At last, the spectral structure of the strange single heavy baryons is shown. Accordingly, we predict several new baryons that might be observed in forthcoming experiments., 27 pages, 11 figures, 8 tables

Additional file1: Supplementary Table 1. Thedifference in basic information between the training and test datasets in MIMIC III and MIMIC IV. Supplementary Table 2. Themodel performance for prediction of in-hospital mortality in the test datasetin MIMIC III and MIMIC IV

In recent years, coal mine has applied shield tunneling machines to roadway excavation to improve production efficiency. Geological condition is an important factor that determines the efficiency of shield machines. The shield machine is most favorable for medium to hard surrounding rocks such as limestone and sandstone. Therefore, the lithology prediction of the location of a planned excavation roadway becomes the core issue in improving the efficiency of the shield machine. At present, seismic inversion is an essential method for lithology prediction. However, in Yangquan Xinjing area, missing P-wave logging curves affects the impedance inversion. Therefore, using existing logging curves to generate missing P-wave logging curves and using sandstone exposure data to continuously update lithology distribution prediction results are of great interest. In this study, logging curves were first pretreated by standardization to ensure the inversion effect. Because of the missing acoustic logging curves, the random forest regression algorithm was introduced using density, natural gamma, apparent resistivity, and spontaneous potential curves as characteristic variables to establish a curve regression prediction model. Then, P-wave logging curves were acquired. After a full analysis of the principles of acoustic and gamma curves, a quasi-acoustic curve is constructed, and a quasi-acoustic inversion was performed. The top and bottom interfaces of the K7 sandstone were interpreted on the inversion data body. The interpreted horizon information was converted from the time domain to the depth domain. The predicted results agreed well with the exposed data. At the same time, combined with the lithology exposure data from the shield tunneling machine, the distribution prediction of the K7 sand body in the target roadway section was updated and iterated many times, which provided effective guidance for the optimization of the tunneling route of the shield tunneling machine.

The cyclic oxidation behavior of Hf/Y-doped B2 FeAl intermetallics at 1373 K was investigated. For an undoped FeAl alloy, premature spallation of the alumina film occurs due to the formation of numerous voids at the film/alloy interface and apparent shrinkage in the film. In contrast to this, doping with either Hf or Y significantly improves the interfacial adhesion between the alumina film and the alloy substrate, particularly with Hf-doping. Microstructural observation in combination with Auger electron spectroscopic analysis suggests that in addition to prohibiting interfacial void formation and alleviating film shrinkage, the addition of Hf in the FeAl alloy could consolidate the film/alloy interface by directly participating in chemical bonding across the interface as a Hf ion. This causes the spallation of alumina film from the equiaxed grains/columnar grains interface rather than the bottom of the film.

In this paper, graphene@cuprous oxide (rGO@Cu2O) nanocomposite was designed and prepared with graphene oxide, CuSO4, NaOH and L-ascorbic acid via an in-situ reaction process, and the as-prepared rGO@Cu2O nanocomposite was characterized by XRD, FT-IR, Raman spectroscopy, XPS, SEM-EDS and TEM. The results reveal that the rGO@Cu2O nanocomposite is of homogeneous distribution, and the Cu2O nanoparticles adsorbed on graphene sheets are with a fairly uniform size of 2.3 nm. The rGO@Cu2O/acrylic resin self-polishing marine antifouling paint with functional surfaces was also prepared in this work, and a series of measurements were carried out for the obtained product. The WCAs of paint is up to 113°, and the adhesive force is averaged to 3.69 MPa. After being immersed into seawater, the whole bared panels show an abundant growth of marine organisms within 90 days, but the rGO@Cu2O paint coated surfaces are hardly fouled by marine organisms within 365 days. This work demonstrates that in situ synthesis of rGO@Cu2O is a tin-free potential alternative to inhibit biofouling.

Compared with other parts of the plant, leaves are easy to obtain and have obvious features. Therefore, leaf identification is also considered as one of the best methods for plant identification. For this reason, a transfer learning leaf recognition method based on VGG16 deep learning network is proposed. This method firstly carries out preprocess operations such as background whitening and standard normalization processing, then carries out data augmentation, then uses transfer learning to train VGG16 network model, and finally completes leaf recognition. The experimental results of Middle European Woody (MEW) plants data set and UCI Folio Leaf data set show that the recognition accuracy under a single background reaches 93.4% and 97.9%, respectively, which are significantly higher than traditional convolutional neural network and other recognition methods, effectively improving the recognition accuracy rate of plant leaves.

With the rapid development of artificial intelligence, the purpose of image classification is not only to identify the major categories of objects, but also to classify the images of the same category into more detailed subcategories. In order to effectively discriminate small differences between regions, a fine-grained classification algorithm was proposed based on Attention Self-supervision (ASS). Firstly, the ResNet-152 pre-trained model was applied to extract the global image features and to forecast the attention maps with Depthwise Separable convolution (DS conv). Subsequently, the Weakly Supervised Data Augmentation Network (WS-DAN) was applied to feed the augmented image back into the network and enhance the generalization ability of the network to prevent overfitting. Finally, linear fusion of attention features in ASS network was further performed to improve the accuracy of the classification. The experimental results show that the method achieves 89.6% accuracy rate on CUB-200-2011 dataset.

Additional file 1: Table S1. Univariate analysis of associations with clinical outcomes. Figure S1. Kaplan-Meier survival plots of mCRPC patients. The survival differences were compared between patients with ≥5 CTCs and those with

A facile method was used to prepare graphene from discarded polyethylene plastic bags in our work. In order to make high-quality graphene, PE plastic bags were ultrasonically cleaned, ball milled and microwave sintered successively. The height of the 2D band was 1.3 times that of the G band, which reveals that the layer number of as-prepared graphene was 1-2. The atomic ratio of C and O for graphene was more than 54, which indicates that it mainly consists of carbon. The size of graphene was within 4-10 μm. Bi-layer sheets were inevitably observed through high resolution imaging of graphene edges. The BET SSA and the electrical conductivity of graphene were 1521.3 m2 g-1 and 4618 S m-1, respectively. This work provides a new approach to large-scale and high-quality synthesis of graphene from waste polluting materials.

Layered adhesive structures are widely used in the aerospace field, and thus their defect detection is a critical part of quality control. Hence, this paper proposes a method to identify bonding defects relying on dynamic time warping (DTW) and simulation analysis. Based on the transfer matrix method, the propagation simulation model of the layered structure is established, and the simulation signal of the normal area and the characteristic interval of the defects are obtained. DTW is adopted to regularize the simulated and the real signals, and then the Pearson correlation coefficient is employed to calculate the regularized signals’ similarity. On this basis, the proposed scheme is considered a similarity imaging method, attaining an appealing identification error for the three- and the five-layered adhesive structure cases examined involving various defect thicknesses and glue layers. Compared to current imaging methods, the proposed one demonstrates a better and more effective imaging effect appropriate for bonding defect detection.

Background Two-line hybrid wheat system using thermo-photo sensitive genic male sterility (TPSGMS) is currently the most promising approach for wheat heterosis utilization in China. However, during past 20 years only few TPSGMS lines were developed in hybrid wheat breeding, which has been the main limiting factor to create heterotic hybrids. Application of doubled haploid (DH) breeding provides a useful strategy to efficiently develop practically usable TPSGMS lines. Results F1s and selected F2 and F3 sterile plants of eight crosses made from two commercial TPSGMS lines were used to produce DH lines. We developed a total of 24 elite DH sterile lines with stable sterility, good outcrossing and yield potential, resistance to yellow rust and powdery mildew, as well as desirable plant height (50–60 cm). These DH lines were developed within 4 years through at least 1 year of evaluation. The stability of male sterility was confirmed for most (20/24) of these elite DH sterile lines by multiple tests in two or 3 years. These lines are expected to be used in hybrid wheat breeding. The percentage of elite lines developed from the tested DH lines produced from filial generations was in the order of F2 > F3 > F1. Conclusions We demonstrate that coupling DH techniques with conventional breeding is an efficient strategy for accelerating the development of more practical wheat TPSGMS lines. Generation of DHs from F2 generation appeared to be the better choice considering the balance of shortening breeding time and overall breeding efficiency.

This paper proposes a knowledge distillation method for foreground object search (FoS). Given a background and a rectangle specifying the foreground location and scale, FoS retrieves compatible foregrounds in a certain category for later image composition. Foregrounds within the same category can be grouped into a small number of patterns. Instances within each pattern are compatible with any query input interchangeably. These instances are referred to as interchangeable foregrounds. We first present a pipeline to build pattern-level FoS dataset containing labels of interchangeable foregrounds. We then establish a benchmark dataset for further training and testing following the pipeline. As for the proposed method, we first train a foreground encoder to learn representations of interchangeable foregrounds. We then train a query encoder to learn query-foreground compatibility following a knowledge distillation framework. It aims to transfer knowledge from interchangeable foregrounds to supervise representation learning of compatibility. The query feature representation is projected to the same latent space as interchangeable foregrounds, enabling very efficient and interpretable instance-level search. Furthermore, pattern-level search is feasible to retrieve more controllable, reasonable and diverse foregrounds. The proposed method outperforms the previous state-of-the-art by 10.42% in absolute difference and 24.06% in relative improvement evaluated by mean average precision (mAP). Extensive experimental results also demonstrate its efficacy from various aspects. The benchmark dataset and code will be release shortly., Comment: This paper will appear at ECCV 2020

Additional file 1 : Table S1. The primers for amplification and sequencing the exons of G6PD gene.

Additional file 1: Table S1. C1-INH antigenic levels of patients with missense mutation data.

Additional file 8. Supplementary Materials and Methods.

In the three-dimensional coordinate system, there are four omnidirectional antenna radar stations with fixed coordinates. These radar stations form a network radar and observe the target. In this paper, the beat signal with white Gaussian noise obtained by mixing the FMCW(Frequency Modulated Continuous Wave) transmission signal and the echo signal is processed by FFT to obtain the frequency value of the beat signal. Finally, the state information including the target distance obtained by the improved Jacobi iterative algorithm is compared and compared with the traditional Jacobi iterative algorithm. The simulation results were analyzed, and the superiority of the improved algorithm was verified, which provided help for accurate and timely meteorological observations.

Anodic aluminum oxide (AAO) thin films were fabricated in phosphoric acid electrolyte solution via one-step electrochemical anodic oxidation of an Al sheet. The films displayed minimally saturated rainbow rings of structural colors, and pores with depths that decreased symmetrically outward from the center of the films. AAO thin films embedded with Co nanodots were fabricated using the one-step ac electrodeposition method. The composite films showed highly saturated rainbow rings of structural colors, which is due to the Co nanodots efficiently screening reflected light from the Al substrate. Moreover, magnetic Co nanodots were found to tune the physical properties of AAO thin film as a new degree of freedom. After determining that Co@AAO composite films possess magnetism and highly saturated rainbow rings of structural colors, their magnetic and optical properties were studied based on modification of the oxidation voltage and deposition time required to prepare the films.

B2 FeAl intermetallic compounds modified with reactive elements (REs) including Sc and Y were fabricated by vacuum arc-melting, and the isothermal oxidation behavior of the RE-doped alloys at 1373 K was investigated. Both Sc and Y single-doping significantly decrease the alumina film growth rate of the alloys. The alumina film growth rate of Sc + Y co-doped alloy even further reduces compared to that of the Sc and Y single-doped alloys. The synergistic effect produced by Sc + Y co-doping on the growth behavior of alumina was discussed. It could be anticipated that the combined additions of Sc and Y which have matched chemical properties might decrease the alumina film growth rate more effectively and provide FeAl alloys with enhanced oxidation resistance.

Hydromulching is a new mulch product which has positive effects on soil moisture conservation and the inhibition of evaporation. However, current research on its soil moisture conservation effect has mainly adopted the field fixed-point observation method, rather than revealing the soil moisture conservation mechanism from the microstructure of the hydromulching. This paper applied spectrum analysis technology to observe the structure and micromorphological features of hydromulching, which will be able to reveal the mechanism of soil moisture conservation. Three types of hydromulching, based on laboratory preparation, were used to analyze the mechanism of soil moisture, using the infrared spectrometer (FTIR) and environmental scanning electron microscope (ESEM). The results showed that −OH generates intermolecular hydrogen bonds between carboxymethylcellulose sodium (CMC) and polyvinyl alcohol (PVA); the hydrogen bonds were formatted between the molecules of the two components, strengthening their compatibility, which increased the effectiveness of the hydromulching and created a dense mesh structure. ESEM observation results showed that preparation 1's hydromulching coverage had good effects, while preparations 2 and 3 performed poorly. Therefore, we can conclude that hydromulching effectively cements soil particles. Then it can generate the soil membrane structure to reduce evaporation and improve soil moisture conservation function.

Development of high-efficiency millimeter-wave-absorbing materials is important to attenuate electromagnetic pollution occurring due to rapid development of information technology and to defend against radar detection for military stealth technology. In this study, BaTiO3 (BTO)/Co3O4 millimeter-wave absorption composite powders with hierarchical microstructures were successfully prepared by the sol–gel method. The results indicated that the Co3O4 phase was preferentially formed at low temperature of 800–900 °C and BTO was formed at high temperature of 1000–1100 °C. A composite with perfect crystallization of both constituent phases and uniform microstructure was obtained when its molar ratio of BTO/Co approached 1 and the sintering temperature was around 900 °C. Both permittivity and permeability resonances were generated typically in composites, due to which er and μr values were relatively closer and e′′ and μ′′ were high at high-frequency range of 30–40 GHz. In the composite with molar ratio of BTO/Co of 4/6 sintered at 900 °C, double absorption peaks with appropriate frequency difference were generated, and absorption intensity from −40 to −50 dB and bandwidth of 5 GHz were successfully obtained with sample matching thickness of 1.8 mm around 35 GHz. It completely covered the atmosphere window at 35 GHz, which has the strongest transparency of millimeter wave. This composite can be used as a highly efficient absorption material of millimeter-wave rays.

AMP-activated protein kinase α1 (AMPK α1) is involved in the tumorigenesis of various cancer types. However, the role of AMPK α1 in non-small cell lung cancer (NSCLC) remains unclear. In the present study, 99 NSCLC tumor tissues and paired normal tissues were obtained. The expression levels of AMPK α1 were significantly upregulated in NSCLC tumor tissues compared with those in adjacent non-tumor lung tissues. The patients were further divided into two groups according to their expression levels of AMPK α1 in tumor tissues. The results outlined that overexpression of AMPK α1 was associated with poor prognosis. In addition, vascular endothelial growth factor (VEGF) expression levels were associated with malignant progression in patients with NSCLC. Patients with NSCLC that overexpressed AMPK α1 and VEGF had the worst outcomes. Moreover, AMPK α1 may positively regulate VEGF expression. These results suggest that AMPK α1 serves a carcinogenic role at least in part through the regulation of VEGF expression, and thus represents a potential treatment target in patients with NSCLC.

Six selected bread wheat double haploid genotypes and eight maize cultivars belonging to four types were used in this study. The six wheat genotypes pollinated with the eight maize cultivars to study the effects of different maize genotypes of distinct types on the wheat haploid embryo ratio. The results showed that the same wheat material when pollinated with different maize cultivar, the embryo rate differed between 1.5 to 3 times and ranging from 15.43% to 47.80%. The average haploid embryo induction rate was highest in sweet type (33.24%) and lowest in super sweet type (29.59%). The wheat genotype A6 recorded the highest embryo rate percentage (34.11%), while A5 recorded the lowest rate (29.36%). The white Sweet Waxy (B4) found to be the highest for embryo rate, followed by the Cloud Sweet Jade 6th (B5) and Jane wo Waxy 1th (B1), while the lowest was recorded by Cloud Super Sweet 2nd (B8) cultivar. The maize genotypes in the same type differed significantly regarding the haploid embryo induction rate of wheat genotypes and thus maize types were not consistent in their behavior. So, the key was to select maize genotypes and not the maize type. The interaction between wheat genotype and maize genotype has a significant effect on the embryogenesis rate. Therefore, if the rate of embryo induction of some wheat materials with a common maize variety is low, it may be possible using the pollination of other maize cultivars to increase the embryogenesis in wheat x maize hybridization.

Table S1. List of Primary Antibodies Used In the study. (DOCX 15 kb)

Table S2. Sequences of Primer for Real-time Polymerase Chain Reaction. (DOCX 13 kb)

Table S1. Sequences of primer for real-time polymerase chain reaction (DOCX 14Â kb)

Table S2. List of primary antibodies used in the study (DOCX 14Â kb)

The accurate detection and imaging of small debonded regions in a 3-layer ceramic matrix composite (CMC) with two bonded layers is a challenge within terahertz (THz) nondestructive detection. To solve this problem, an improved imaging method for detecting bonding defects is proposed based on the statistical characteristics of variance and kurtosis to detect defects in the upper/lower adhesive layers in a multilayered CMC. The proposed method can ultimately realize a detection accuracy of 50 and 250 μm for bonding defects in the upper and lower adhesive layers, respectively, and achieve a better imaging result compared with power imaging method. Moreover, there is a corresponding relationship between the statistical characteristic value and the bonding defect size—the larger the statistical characteristic value, the larger the bonding defect size—which can be used as a reference to quantify the defect size.

We propose a THz nondestructive analysis method based on multiple echoes and feature fusion. Conventionally, it is difficult to identify the debonding defects of the glue layer (II) due to the thin adhesive layer. To this end, a THz propagation model is established, and a quantitative method for determining the thickness of debonding defects based on multiple echoes is presented. The measurement error for preset defect thickness of 500 µm was 4%. Further, for determining the area of debonding defects, a feature fusion imaging algorithm is proposed to realize the lateral recognition of defects and quantitative analysis is used to improve the recognition ability of defects.

In order to improve and solve the capability of small and medium enterprises to design new products, the principle of operation of the three-lobe Roots blower rotor was firstly analyzed, and the characteristics of meshing of the rotor profile were analyzed. The meshing design formula was given. Based on this, the 3L41WD involute Roots blower design was taken as an example to calculate its full line type and to analyze and determine the meshing tolerance. Finally, the field test proved the rationality of the design, and provided guidance for SMEs to rapidly design the involute Roots blower rotor.