Xi'an Jiaotong University

About: Xi'an Jiaotong University is a education organization based out in Xi'an, China. It is known for research contribution in the topics: Heat transfer & Dielectric. The organization has 85440 authors who have published 99682 publications receiving 1579683 citations. The organization is also known as: '''Xi'an Jiaotong University''' & Xi'an Jiao Tong University.

The climatic cyclicity in semiarid‐arid central Asia over the past 500,000 years

Abstract: [1] Central Asia is currently a semiarid-arid region, dominated by the Westerlies. It is important to understand mechanisms of climate and precipitation changes here, as water availability in the region is crucial today and in the future. High-resolution, absolutely-dated oxygen isotope (δ18O) records of stalagmites from Kesang Cave characterize a dynamic precipitation history over most of the past 500,000 years. This record demonstrates, for the first time, that climate change in the region exhibits a processional rhythm with abrupt inceptions of low δ18O speleothem growth at times of high Northern Hemisphere summer insolation followed by gradual δ18O increases that track decreases of insolation. These observations and interpretations contrast with the interpretation of nearby, but higher elevation ice core records. The absolutely-dated caveδ18O shifts can be used to correlate the regional climate variability by providing chronological marks. Combined with other paleoclimate records, the Kesang observations suggest that possible incursions of Asian summer monsoon rainfall or related moisture into the Kesang site and/or adjacent areas during the high insolation times may play an important role in changing orbital-scale hydrology of the region. Based on our record, arid climate will prevail in this region for the next several millennia, providing that anthropogenic effects do not supersede natural processes.

Encapsulated Metal Nanoparticles for Catalysis.

Abstract: Metal nanoparticles have drawn great attention in heterogeneous catalysis One challenge is that they are easily deactivated by migration-coalescence during the catalysis process because of their high surface energy With the rapid development of nanoscience, encapsulating metal nanoparticles in nanoshells or nanopores becomes one of the most promising strategies to overcome the stability issue of the metal nanoparticles Besides, the activity and selectivity could be simultaneously enhanced by taking advantage of the synergy between the metal nanoparticles and the encapsulating materials as well as the molecular sieving property of the encapsulating materials In this review, we provide a comprehensive summary of the recent progress in the synthesis and catalytic properties of the encapsulated metal nanoparticles This review begins with an introduction to the synthetic strategies for encapsulating metal nanoparticles with different architectures developed to date, including their encapsulation in nanoshells of inorganic oxides and carbon, porous materials (zeolites, metal-organic frameworks, and covalent organic frameworks), and organic capsules (dendrimers and organic cages) The advantages of the encapsulated metal nanoparticles are then discussed, such as enhanced stability and recyclability, improved selectivity, strong metal-support interactions, and the capability of enabling tandem catalysis, followed by the introduction of some representative applications of the encapsulated metal nanoparticles in thermo-, photo-, and electrocatalysis At the end of this review, we discuss the remaining challenges associated with the encapsulated metal nanoparticles and provide our perspectives on the future development of the field

Bayesian Loss for Crowd Count Estimation With Point Supervision

Abstract: In crowd counting datasets, each person is annotated by a point, which is usually the center of the head. And the task is to estimate the total count in a crowd scene. Most of the state-of-the-art methods are based on density map estimation, which convert the sparse point annotations into a “ground truth” density map through a Gaussian kernel, and then use it as the learning target to train a density map estimator. However, such a "ground-truth" density map is imperfect due to occlusions, perspective effects, variations in object shapes, etc. On the contrary, we propose Bayesian loss, a novel loss function which constructs a density contribution probability model from the point annotations. Instead of constraining the value at every pixel in the density map, the proposed training loss adopts a more reliable supervision on the count expectation at each annotated point. Without bells and whistles, the loss function makes substantial improvements over the baseline loss on all tested datasets. Moreover, our proposed loss function equipped with a standard backbone network, without using any external detectors or multi-scale architectures, plays favourably against the state of the arts. Our method outperforms previous best approaches by a large margin on the latest and largest UCF-QNRF dataset.

Deep Convolution Neural Networks for Twitter Sentiment Analysis

Abstract: Twitter sentiment analysis technology provides the methods to survey public emotion about the events or products related to them. Most of the current researches are focusing on obtaining sentiment features by analyzing lexical and syntactic features. These features are expressed explicitly through sentiment words, emoticons, exclamation marks, and so on. In this paper, we introduce a word embeddings method obtained by unsupervised learning based on large twitter corpora, this method using latent contextual semantic relationships and co-occurrence statistical characteristics between words in tweets. These word embeddings are combined with n-grams features and word sentiment polarity score features to form a sentiment feature set of tweets. The feature set is integrated into a deep convolution neural network for training and predicting sentiment classification labels. We experimentally compare the performance of our model with the baseline model that is a word n-grams model on five Twitter data sets, the results indicate that our model performs better on the accuracy and F1-measure for twitter sentiment classification.

Co-Ln mixed-metal phosphonate grids and cages as molecular magnetic refrigerants

Abstract: The synthesis, structures, and magnetic properties of six families of cobalt–lanthanide mixed-metal phosphonate complexes are reported in this Article. These six families can be divided into two structural types: grids, where the metal centers lie in a single plane, and cages. The grids include [4 × 3] {Co8Ln4}, [3 × 3] {Co4Ln6}, and [2 × 2] {Co4Ln2} families and a [4 × 4] {Co8Ln8} family where the central 2 × 2 square is rotated with respect to the external square. The cages include {Co6Ln8} and {Co8Ln2} families. Magnetic studies have been performed for these compounds, and for each family, the maximum magnetocaloric effect (MCE) has been observed for the Ln = Gd derivative, with a smaller MCE for the compounds containing magnetically anisotropic 4f-ions. The resulting entropy changes of the gadolinium derivatives are (for 3 K and 7 T) 11.8 J kg–1 K–1 for {Co8Gd2}; 20.0 J kg–1 K–1 for {Co4Gd2}; 21.1 J kg–1 K–1 for {Co8Gd4}; 21.4 J kg–1 K–1 for {Co8Gd8}; 23.6 J kg–1 K–1 for {Co4Gd6}; and 28.6 J kg–1 K–1 ...