Sign language production aims to automatically generate coordinated sign language videos from spoken language. As a typical sequence to sequence task, the existing methods are mostly to regard the ...skeletons as a whole sequence, however, those do not take the rich graph information among both joints and edges into consideration. In this paper, we propose a novel method named Spatial-Temporal Graph Transformer (STGT) to deal with this problem. Specifically, according to kinesiology, we first design a novel graph representation to achieve graph features from skeletons. Then the spatial-temporal graph self-attention utilizes graph topology to capture the intra-frame and inter-frame correlations, respectively. Our key innovation is that the attention maps are calculated on both spatial and temporal dimensions in turn, meanwhile, graph convolution is used to strengthen the short-term features of skeletal structure. Finally, due to the generated skeletons are based on the form of skeleton points and lines so far. In order to visualize the generated sign language videos, we design a sign mesh regression module to render the skeletons into skinned animations including body and hands posture. Comparing with states of art baseline on RWTH-PHONEIX Weather-2014T in Experiment Section, STGT can obtain the highest values on BLEU and ROUGE, which indicates our method produces most accurate and intuitive sign language videos.
Autophagy during myocardial ischemia/reperfusion (MI/R) exacerbates cardiomyocyte injury. Melatonin (Mel) alleviates myocardial damage by regulating mitochondrial function and mitophagy, but the role ...of mitophagy in melatonin-induced cardioprotection remains unclear. This study aimed to explore the roles of sirtuin3 (SIRT3) and retinoid‐related orphan nuclear receptor‐α (RORα) in mitophagy during simulated ischemia reperfusion (SIR) in H9c2 cells. Our data showed that mitophagy was excessively activated after SIR injury, which was consistent with reduced cell survival, enhanced oxidative responses and mitochondrial dysfunction in H9c2 myocytes. Melatonin greatly enhanced cell viability, reduced oxidative stress and improved mitochondrial function. The effects of melatonin protection were involved in excessive mitophagy inhibition, as demonstrated by the reduced levels of mitophagy-linked proteins, including Parkin, Beclin1, NIX and BNIP3, and the LC3 II/LC3 I ratio and elevations in p62. Additionally, the decreases in SIRT3 and RORα in H9c2 myocytes after SIR were reversed by melatonin, and the above effects of melatonin were eliminated by small interfering RNA (siRNA)-mediated knockdown of SIRT3 and RORα. In brief, SIRT3 and RORα are two prospective targets in the cardioprotection of melatonin against mitophagy during SIR in H9c2 myocytes.
With the goal of achieving carbon neutrality, active distribution networks (DNs) with a high proportion of photovoltaics (PVs) are facing challenges in maintaining voltage stability and low‐carbon ...operation. Energy storage systems (ESSs), which have the ability to store and transfer energy temporarily, can be used as effective measures to enhance the capacity of consuming PVs and reduce carbon emissions in DNs. However, existing low‐carbon dispatch strategies for multiple sources, storages and loads fail to consider voltage violations, while the temporal carbon emission intensity of the upper‐level power grid is also often overlooked, which is an important factor that affects the dispatch strategy. Therefore, a two‐stage self‐adaptive dispatch strategy of ESSs that considers the temporal characteristics of slack nodal carbon emission intensity to minimise carbon emissions while maintaining voltage stability in DNs with high access to PVs is proposed. First, the framework of the proposed two‐stage self‐adaptive dispatch strategy of ESSs is established by taking into account the effects of ESSs on adjusting voltages and reducing carbon emissions, respectively, with the two‐stage switch principle of two operation modes being determined. On this basis, an optimization dispatch model is established to improve voltages and carbon emissions, and the optimal day‐ahead dispatch strategy of ESSs can be obtained by solving the model using genetic algorithm. Case studies of the modified 10 kV IEEE 33‐node DN and IEEE 123‐node DN verify the feasibility and superiority of the proposed two‐stage self‐adaptive security and low‐carbon day‐ahead dispatch strategy for ESSs, showing that the voltage stabilisation and lower carbon emissions of DNs are both improved.
Active distribution networks with a high proportion of photovoltaics (PVs) are facing challenges in maintaining voltage stability and low‐carbon operation. The feasibility and superiority of the proposed two‐stage self‐adaptive security and low‐carbon day‐ahead dispatch strategy for ESSs is verified, showing that the voltage stabilisation and lower carbon emissions of distribution networks are both improved.
Objectives
To evaluate the diagnostic accuracy of unenhanced and contrast-enhanced MRI in the differentiation of mucosal high-grade neoplasia (MHN) from early invasive squamous cell cancer (EISCC) of ...the esophagus.
Methods
Between March 2015 and January 2019, 72 study participants with MHN (
n
= 46) and EISCC (
n
= 26) of the esophagus were enrolled in this prospective study. Postoperative histopathologic analysis was the reference standard. All participants underwent MRI (T2-multi-shot turbo spin-echo sequence (msTSE), diffusion-weighted imaging (DWI), and 3D gradient-echo-based sequence (3D-GRE)). Two radiologists, blinded to participants’ data, independently evaluated MRI and assigned MR features including shape (mucosal thickening or focal mass), signal on T2-msTSE and DWI, enhancement degree (intense or slight), and enhancement pattern (homogeneous, heterogeneous, or heart-shaped). Diagnostic performance of the 5 features was compared using the chi-square test; kappa values were assessed for reader performance.
Results
Surgery was performed within 3.6 + 3.5 days after MR imaging. Inter-reader agreement on MR features was excellent (kappa value = 0.854,
p
< 0.001). All 8 mass-like MHN were “heart-shaped” in appearance. The degree of enhancement showed the best diagnosis performance in differentiating between MHN and EISCC of the esophagus. The combination of all 5 features had only borderline improved sensitivity, specificity, and AUC of 100%, 96.2%, and 0.999, respectively, which was not statistically significant compared with the degree of enhancement alone.
Conclusions
MRI can differentiate MHN from EISCC in esophagus; the presence of “heart-shaped” appearance favors the diagnosis of MHN.
Key Points
• All 8 mass-like MHN showed a “heart-shaped” enhancement pattern which may help differentiating MHN from EISCC.
• Degree of enhancement had the best diagnostic performance in differentiating between MHN and EISCC in esophagus.
• The combined 5 features (shape, signal in T2-msTSE and DWI, enhancement degree, and enhancement pattern) provided sensitivity, specificity, and AUC of 100%, 96.2%, and 0.999, respectively, which was not statistically significant than tumor enhancement alone in distinguishing MHN from EISCC.
We propose a method for simulating cloth with meshes dynamically refined according to visual saliency. It is a common belief that it is preferable for the regions of an image being viewed to have ...more details than others. For a certain scene, a low-resolution cloth mesh is first simulated and rendered into images in the preview stage. Pixel saliency values of these images are predicted according to a pre-trained saliency prediction model. These pixel saliencies are then translated to a vertex saliency of the corresponding meshes. Vertex saliency, together with camera positions and a number of geometric features of surfaces, guides the dynamic remeshing for simulation in the production stage. To build the saliency prediction model, images extracted from various videos of clothing scenes were used as training data. Participants were asked to watch these videos and their eye motion was tracked. A saliency map is generated from the eye motion data for each extracted video frame image. Image feature vectors and map labels are sent to a Support Vector Machine for training to obtain a saliency prediction model. Our method greatly reduces the number of vertices and faces in the clothing model, and generates a speed-up of more than 3 × for scenes with single dressed character, while for multi-character scenes the speed-up is increased to more than 5×. The proposed technique can work together with view-dependency for offline simulation.
Sulfonated poly(phthalazinone ether ketones) (SPPEK) with ion exchange capacities from 0.77 to 1.82 mmol·g
are synthesized via an electrophilic substitution reaction. Nuclear magnetic resonance and ...infrared absorption spectroscopy are used to characterize the chemical structure of the obtained polymers for confirming the successful introduction of sulfonic groups. SPPEKs show excellent thermal stability; their temperature required to achieve 5% weight loss is about 360 °C. Accordingly, the obtained membranes possess high ion perm-selectivity, proton conductivity, and low area resistance. Regarding the electrodialysis-related performance of the membranes, the SPPEK-4 membrane has the highest limiting current density (39.8 mA·cm
), resulting from its high content of sulfonic groups. In a desalination test of standard solution, SPPEK-3 and SPPEK-4 membranes exhibit both better salt removal rate and acceptable energy consumption than commercial membrane. Additionally, SPPEK-3 membrane shows outstanding performance in terms of high concentration rate and low energy consumption during saline water treatment, which indicates the feasibility of novel membranes in electrodialysis application.
The Three-River Source Region (TRSR) is vital to the ecological security of China. However, the impact of global warming on the dynamics of vegetation along the elevation gradient in the TRSR remains ...unclear. Accordingly, we used multi-source remote sensing vegetation indices (VIs) (GIMMS (Global Inventory Modeling and Mapping Studies) LAI (Leaf Area Index), GIMMS NDVI (Normalized Difference Vegetation Index), GLOBMAP (Global Mapping) LAI, MODIS (Moderate Resolution Imaging Spectroradiometer) EVI (Enhanced Vegetation Index), MODIS NDVI, and MODIS NIRv (near-infrared reflectance of vegetation)) and digital elevation model data to study the changes of VGEG (Vegetation Greenness along the Elevation Gradient) in the TRSR from 2001 to 2016. Results showed that the areas with a positive correlation of vegetation greenness and elevation accounted for 36.34 ± 5.82% of the study areas. The interannual variations of VGEG showed that the significantly changed regions were mainly observed in the elevation gradient of 4–5 km. The VGEG was strongest in the elevation gradient of 4–5 km and weakest in the elevation gradient of >5 km. Correlation analysis showed that the mean annual temperature was positively correlated with VIs, and the effect of the mean annual precipitation on VIs was more obvious at low altitude than in high altitude. This study contributes to our understanding of the VGEG variation in the TRSR under global climate variation and also helps in the prediction of future carbon cycle patterns.
Depth estimation is a classical problem in computer vision, which typically relies on either a depth sensor or stereo matching alone. The depth sensor provides real-time estimates in repetitive and ...textureless regions where stereo matching is not effective. However, stereo matching can obtain more accurate results in rich texture regions and object boundaries where the depth sensor often fails. We fuse stereo matching and the depth sensor using their complementary characteristics to improve the depth estimation. Here, texture information is incorporated as a constraint to restrict the pixel's scope of potential disparities and to reduce noise in repetitive and textureless regions. Furthermore, a novel pseudo-two-layer model is used to represent the relationship between disparities in different pixels and segments. It is more robust to luminance variation by treating information obtained from a depth sensor as prior knowledge. Segmentation is viewed as a soft constraint to reduce ambiguities caused by under- or over-segmentation. Compared to the average error rate 3.27% of the previous state-of-the-art methods, our method provides an average error rate of 2.61% on the Middlebury datasets, which shows that our method performs almost 20% better than other "fused" algorithms in the aspect of precision.
We introduce a novel approach for flame volume reconstruction from videos using inexpensive charge-coupled device (CCD) consumer cameras. The approach includes an economical data capture technique ...using inexpensive CCD cameras. Leveraging the smear feature of the CCD chip, we present a technique for synchronizing CCD cameras while capturing flame videos from different views. Our reconstruction is based on the radiative transport equation which enables complex phenomena such as emission, extinction, and scattering to be used in the rendering process. Both the color intensity and temperature reconstructions are implemented using the CUDA parallel computing framework, which provides real-time performance and allows visualization of reconstruction results after every iteration. We present the results of our approach using real captured data and physically-based simulated data. Finally, we also compare our approach against the other state-of-the-art flame volume reconstruction methods and demonstrate the efficacy and efficiency of our approach in four different applications: (1) rendering of reconstructed flames in virtual environments, (2) rendering of reconstructed flames in augmented reality, (3) flame stylization, and (4) reconstruction of other semitransparent phenomena.
As a key factor that determines carbon storage capacity, residence time (τE) is not well constrained in terrestrial biosphere models. This factor is recognized as an important source of model ...uncertainty. In this study, to understand how τE influences terrestrial carbon storage prediction in diagnostic models, we introduced a model decomposition scheme in the Boreal Ecosystem Productivity Simulator (BEPS) and then compared it with a prognostic model. The result showed that τE ranged from 32.7 to 158.2 years. The baseline residence time (τ'E) was stable for each biome, ranging from 12 to 53.7 years for forest biomes and 4.2 to 5.3 years for non-forest biomes. The spatiotemporal variations in τE were mainly determined by the environmental scalar (ξ). By comparing models, we found that the BEPS uses a more detailed pool construction but rougher parameterization for carbon allocation and decomposition. With respect to ξ comparison, the global difference in the temperature scalar (ξt) averaged 0.045, whereas the moisture scalar (ξw) had a much larger variation, with an average of 0.312. We propose that further evaluations and improvements in τ'E and ξw predictions are essential to reduce the uncertainties in predicting carbon storage by the BEPS and similar diagnostic models.
PDF
