Search

In the context of global climate change, ocean iron fertilization (OIF) has been suggested as a potential geoengineering strategy to enhance the growth of marine phytoplankton, subsequently promote the ocean carbon sink, and ultimately regulate atmospheric carbon dioxide (CO2) and mitigate climate change. However, in past artificial OIF experiments, the elevation of both net primary production and carbon export to the deep ocean is far less than anticipated. This discrepancy can be attributed to several key factors that strongly influence the efficiency of OIF and biological carbon pump (BCP) including iron bioavailability and retention time, the sensitivity of phytoplankton to iron limitation, the influence of viruses and other microorganisms comprising the microbial carbon pump (MCP), and non-biological carbon cycles. Although previous reviews have extensively summarized the outcomes of the past OIF experiments, the mechanisms and key factors that influence climate regulation mediated by OIF have not been elucidated. In this review, we analyze these mechanisms throughout the whole process from OIF, to phytoplankton physiological responses, to carbon dioxide removal (CDR) and sequestration in the ocean, and ultimately to CDR in the atmosphere and its impact on climate change. By examining both natural and artificial OIF cases, as well as recent advancements in phytoplankton physiological ecology and carbon pump theory, possible reasons leading to the inconsistencies between OIF and carbon export are explored. These comprehensive analyses are helpful to identify the key issues in regulating climate change by providing a reference and scientific approach to guide potential OIF attempts in the future. © 2023

Yang Zhao,1,2 Xu Tong Yang,1 Yan Ping Bai,3 Lin Feng Li2 1Department of Dermatology, Daxing Teaching Hospital, Capital Medical University, Beijing, 102600, People’s Republic of China; 2Department of Dermatology, Beijing Friendship Hospital, Capital Medical University, Beijing, 100050, People’s Republic of China; 3Department of Dermatology, China-Japan Friendship Hospital, Beijing, 100029, People’s Republic of ChinaCorrespondence: Lin Feng Li, Department of Dermatology, Beijing Friendship Hospital, Capital Medical University, 95 Yong an Road, Xicheng District, Beijing, 100050, People’s Republic of China, Email zoonli@sina.comBackground and Aim: Psoriasis is a persistent inflammatory disorder that affects 3% of the population and is associated with cardiovascular diseases. Therefore, this study aimed to investigate the correlations between complete blood cell count (CBC)-derived inflammatory biomarkers, psoriasis prevalence, and all-cause mortality.Methods: Data were obtained from the National Health and Nutrition Examination Survey (NHANES) spanning 2003– 2006 and 2009– 2014. Mortality data up to December 31, 2019 were obtained using the National Death Index. The following CBC-derived inflammatory biomarkers were examined: neutrophil/lymphocyte ratio (NLR), neutrophil/(white blood cells ˗ neutrophils) ratio (dNLR), monocyte count/lymphocyte ratio (MLR), (neutrophil + monocyte)/lymphocyte ratio (NMLR), platelet/lymphocyte ratio (PLR), and systemic inflammatory response index (SIRI). Weighted logistic and Cox regression analyses were used to calculate odds ratios, hazard ratios, and the corresponding 95% confidence intervals. Additionally, the prognostic value of the inflammatory indicators was assessed and ranked using the random survival forest approach.Results: In total, data from 21,431 participants with average age of 45.02 ± 0.27 years (49.51% male) were included in the study, among which 600 participants were positive for psoriasis (prevalence rate

Xinghua Guo,1,2,* Pan Song,1,2,* Feng Li,1,2,* Qihao Yan,1,2,* Yan Bai,3 Jincan He,3 Qishi Che,4 Hua Cao,5 Jiao Guo,2 Zhengquan Su1 1Guangdong Engineering Research Center of Natural Products and New Drugs, Guangdong Provincial University Engineering Technology Research Center of Natural Products and Drugs, Guangdong Pharmaceutical University, Guangzhou, 510006, People’s Republic of China; 2Guangdong Metabolic Disease Research Center of Integrated Chinese and Western Medicine, Key Laboratory of Glucolipid Metabolic Disorder, Ministry of Education of China, Guangdong TCM Key Laboratory for Metabolic Diseases, Guangdong Pharmaceutical University, Guangzhou, 510006, People’s Republic of China; 3School of Public Health, Guangdong Pharmaceutical University, Guangzhou, 510310, People’s Republic of China; 4Guangzhou Rainhome Pharm & Tech Co., Ltd, Science City, Guangzhou, 510663, People’s Republic of China; 5School of Chemistry and Chemical Engineering, Guangdong Pharmaceutical University, Zhongshan, 528458, People’s Republic of China*These authors contributed equally to this workCorrespondence: Zhengquan Su, Guangdong Engineering Research Center of Natural Products and New Drugs, Guangdong Pharmaceutical University, Guangzhou, 510006, People’s Republic of China, Tel +86-20-3935-2067, Fax +86-20-3935-2067, Email suzhq@scnu.edu.cnAbstract: Bone, like most organs, has the ability to heal naturally and can be repaired slowly when it is slightly injured. However, in the case of bone defects caused by diseases or large shocks, surgical intervention and treatment of bone substitutes are needed, and drugs are actively matched to promote osteogenesis or prevent infection. Oral administration or injection for systemic therapy is a common way of administration in clinic, although it is not suitable for the long treatment cycle of bone tissue, and the drugs cannot exert the greatest effect or even produce toxic and side effects.

The development of clinical imaging techniques significantly improves diagnostic accuracy and provides guidance for personalized treatment of individuals. However, every single imaging modality has its distinct drawbacks that cannot fully fulfil the diagnosis requirement. Thus, rational combination of different imaging modalities can achieve more comprehensive information of disease and in this way provide better personalized treatment strategy. The hybrid PET/MRI has drawn increasing attention since its first clinical application. Imaging probes play an essential role in achieving qualified figures with accurate information of diseases. The application of nanotechnology promotes the development of versatile molecular probes for PET/MRI technique. Though there is an emerging clinical requirement, only a small number of multimodal PET/MRI probes have been investigated in preclinical research. Thus, this review tries to thoroughly summarize the nano-sized PET/MRI probes on their design, preparation and biological application. By discussing the strength and limitations of these current available PET/MRI multimodal probes, this work aims to figure out the further research direction and promote the possible clinic translation of the novel PET/MRI probes., Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., RST/Applied Radiation & Isotopes

The thermodynamics of the BTZ black holes are revisited with variable Newton constant. A new pair of conjugated variables, the central charge $C$ and the chemical potential $\mu$, is introduced as thermodynamic variables. The first law of thermodynamics and the Euler relation, instead of the Smarr relation in the extended phase space formalism, are matched perfectly in this formalism. Compatible with standard extensive thermodynamics, the black hole mass is verified to be a first order homogeneous function of the related extensive variables, and restores the role of internal energy. In addition, the heat capacity has also resulted in a first order homogeneous function in this formalism as we expected, and an asymptotic behavior in high temperature limit is shown intriguingly. The non-negatively of heat capacity indicates that the rotating and charged BTZ black holes are thermodynamically stable., Comment: 16 pages, no figure

Seawater pH is a direct proxy of ocean acidification, and monitoring the global pH distribution and long-term series changes is critical to understanding the changes and responses of the marine ecology and environment under climate change. Owing to the lack of sufficient global-scale pH data and the complex relationship between seawater pH and related environmental variables, generating time-series products of satellite-derived global sea surface pH poses a great challenge. In this study, we solved the problem of the lack of sufficient data for pH algorithm development by using the massive underway sea surface carbon dioxide partial pressure (pCO(2)) dataset to structure a large data volume of near in situ pH based on carbonate calculation between underway pCO(2) and calculated total alkalinity from sea surface salinity and relevant parameters. The remote sensing inversion model of pH was then constructed through this massive pH training dataset and machine learning methods. After several tests of machine learning methods and groups of input parameters, we chose the random forest model with longitude, latitude, sea surface temperature (SST), chlorophyll a (Chla), and Mixed layer depth (MLD) as model inputs with the best performance of correlation coefficient (R-2 = 0.96) and root mean squared error (RMSE = 0.008) in the training set and R-2 = 0.83 (RMSE = 0.017) in the testing set. The sensitivity analysis of the error variation induced by the uncertainty of SST and Chla (SST <= +/- 0.5 degrees C and Chla <= +/- 20%; RMSESST <= 0.011 and RMSEChla <= 0.009) indicated that our sea surface pH model had good robustness. Monthly average global sea surface pH products from 2004 to 2019 with a spatial resolution of 0.25 degrees x 0.25 degrees were produced based on the satellite-derived SST and Chla products and modeled MLD dataset. The pH model and products were validated using another independent station-measured pH dataset from the Global Ocean Data Analysis Project (GL

Chun-Shuang Guan,1,&ast; Zhi-Juan Liu,2,&ast; Yan-Ni Du,1,&ast; Hui Chen,1 Yan Bai,1 Zhi-Bin Lv,1 Yan-Li Xu,2 Ru-Ming Xie,1,&ast; Bu-Dong Chen1,&ast; 1Department of Radiology, Beijing Ditan Hospital, Capital Medical University, Beijing, People’s Republic of China; 2Chronic Disease Management Centre, Beijing Ditan Hospital, Capital Medical University, Beijing, People’s Republic of China&ast;These authors contributed equally to this workCorrespondence: Ru-Ming Xie; Bu-Dong Chen, Tel +8613911320739 ; +8613801358954, Email mingrux1@163.com; chenbudong00@126.comPurpose: To retrospectively analyse the CT imaging during the long-term follow-up of COVID-19 patients after discharge.Patients and Methods: A total of 122 patients entered the study group. All patients underwent CT examinations. The CT images, which included distribution and imaging signs, were evaluated by two chest radiologists. Laboratory examinations included routine blood work, biochemical testing, and SARS-CoV-2 antibody screening. Statistical methods include chi-square, Fisher’s exact test, one-way analysis of variance, rank sum test and logistic regression by SPSS 17.0.Results: There were 22 (18.0%) patients in the mild group, 74 (60.7%) patients in the moderate group, and 26 (21.3%) patients in the severe–critical group. The median follow-up interval was 405 days (378.0 days, 462.8 days). Only monocytes, prothrombin activity, and γ-glutamyltransferase showed significant differences among the three groups. We found that the more severe the patient’s condition, the more SARS-CoV-2 IgG antibodies existed. Only 11 patients (11.0%) showed residual lesions on CT. The CT manifestations included irregular linear opacities in nine cases (9.0%), reticular patterns in six cases (6.0%), and GGOs in five cases (5.0%).Conclusion: The proportion of residual lesions on CT in COVID-19 patients was significantly reduced after long-term follow-up. The patients’ age and disease condition

Seawater pH is a direct proxy of ocean acidification, and monitoring the global pH distribution and long-term series changes is critical to understanding the changes and responses of the marine ecology and environment under climate change. Owing to the lack of sufficient global-scale pH data and the complex relationship between seawater pH and related environmental variables, generating time-series products of satellite-derived global sea surface pH poses a great challenge. In this study, we solved the problem of the lack of sufficient data for pH algorithm development by using the massive underway sea surface carbon dioxide partial pressure (pCO(2)) dataset to structure a large data volume of near in situ pH based on carbonate calculation between underway pCO(2) and calculated total alkalinity from sea surface salinity and relevant parameters. The remote sensing inversion model of pH was then constructed through this massive pH training dataset and machine learning methods. After several tests of machine learning methods and groups of input parameters, we chose the random forest model with longitude, latitude, sea surface temperature (SST), chlorophyll a (Chla), and Mixed layer depth (MLD) as model inputs with the best performance of correlation coefficient (R-2 = 0.96) and root mean squared error (RMSE = 0.008) in the training set and R-2 = 0.83 (RMSE = 0.017) in the testing set. The sensitivity analysis of the error variation induced by the uncertainty of SST and Chla (SST <= +/- 0.5 degrees C and Chla <= +/- 20%; RMSESST <= 0.011 and RMSEChla <= 0.009) indicated that our sea surface pH model had good robustness. Monthly average global sea surface pH products from 2004 to 2019 with a spatial resolution of 0.25 degrees x 0.25 degrees were produced based on the satellite-derived SST and Chla products and modeled MLD dataset. The pH model and products were validated using another independent station-measured pH dataset from the Global Ocean Data Analysis Project (GL

Continual Test-Time Adaptation (CTTA) aims to adapt the source model to continually changing unlabeled target domains without access to the source data. Existing methods mainly focus on model-based adaptation in a self-training manner, such as predicting pseudo labels for new domain datasets. Since pseudo labels are noisy and unreliable, these methods suffer from catastrophic forgetting and error accumulation when dealing with dynamic data distributions. Motivated by the prompt learning in NLP, in this paper, we propose to learn an image-level visual domain prompt for target domains while having the source model parameters frozen. During testing, the changing target datasets can be adapted to the source model by reformulating the input data with the learned visual prompts. Specifically, we devise two types of prompts, i.e., domains-specific prompts and domains-agnostic prompts, to extract current domain knowledge and maintain the domain-shared knowledge in the continual adaptation. Furthermore, we design a homeostasis-based prompt adaptation strategy to suppress domain-sensitive parameters in domain-invariant prompts to learn domain-shared knowledge more effectively. This transition from the model-dependent paradigm to the model-free one enables us to bypass the catastrophic forgetting and error accumulation problems. Experiments show that our proposed method achieves significant performance gains over state-of-the-art methods on four widely-used benchmarks, including CIFAR-10C, CIFAR-100C, ImageNet-C, and VLCS datasets., Comment: AAAI2023 Outstanding Student Paper Award

Few-Shot Classification(FSC) aims to generalize from base classes to novel classes given very limited labeled samples, which is an important step on the path toward human-like machine learning. State-of-the-art solutions involve learning to find a good metric and representation space to compute the distance between samples. Despite the promising accuracy performance, how to model uncertainty for metric-based FSC methods effectively is still a challenge. To model uncertainty, We place a distribution over class probability based on the theory of evidence. As a result, uncertainty modeling and metric learning can be decoupled. To reduce the uncertainty of classification, we propose a Bayesian evidence fusion theorem. Given observed samples, the network learns to get posterior distribution parameters given the prior parameters produced by the pre-trained network. Detailed gradient analysis shows that our method provides a smooth optimization target and can capture the uncertainty. The proposed method is agnostic to metric learning strategies and can be implemented as a plug-and-play module. We integrate our method into several newest FSC methods and demonstrate the improved accuracy and uncertainty quantification on standard FSC benchmarks., Comment: arXiv admin note: text overlap with arXiv:2107.10161 by other authors

Despite remarkable efforts been made, the classification of gigapixels whole-slide image (WSI) is severely restrained from either the constrained computing resources for the whole slides, or limited utilizing of the knowledge from different scales. Moreover, most of the previous attempts lacked of the ability of uncertainty estimation. Generally, the pathologists often jointly analyze WSI from the different magnifications. If the pathologists are uncertain by using single magnification, then they will change the magnification repeatedly to discover various features of the tissues. Motivated by the diagnose process of the pathologists, in this paper, we propose a trusted multi-scale classification framework for the WSI. Leveraging the Vision Transformer as the backbone for multi branches, our framework can jointly classification modeling, estimating the uncertainty of each magnification of a microscope and integrate the evidence from different magnification. Moreover, to exploit discriminative patches from WSIs and reduce the requirement for computation resources, we propose a novel patch selection schema using attention rollout and non-maximum suppression. To empirically investigate the effectiveness of our approach, empirical experiments are conducted on our WSI classification tasks, using two benchmark databases. The obtained results suggest that the trusted framework can significantly improve the WSI classification performance compared with the state-of-the-art methods.

Few-shot class-incremental learning(FSCIL) focuses on designing learning algorithms that can continually learn a sequence of new tasks from a few samples without forgetting old ones. The difficulties are that training on a sequence of limited data from new tasks leads to severe overfitting issues and causes the well-known catastrophic forgetting problem. Existing researches mainly utilize the image information, such as storing the image knowledge of previous tasks or limiting classifiers updating. However, they ignore analyzing the informative and less noisy text information of class labels. In this work, we propose leveraging the label-text information by adopting the memory prompt. The memory prompt can learn new data sequentially, and meanwhile store the previous knowledge. Furthermore, to optimize the memory prompt without undermining the stored knowledge, we propose a stimulation-based training strategy. It optimizes the memory prompt depending on the image embedding stimulation, which is the distribution of the image embedding elements. Experiments show that our proposed method outperforms all prior state-of-the-art approaches, significantly mitigating the catastrophic forgetting and overfitting problems.

Real-world visual search systems involve deployments on multiple platforms with different computing and storage resources. Deploying a unified model that suits the minimal-constrain platforms leads to limited accuracy. It is expected to deploy models with different capacities adapting to the resource constraints, which requires features extracted by these models to be aligned in the metric space. The method to achieve feature alignments is called ``compatible learning''. Existing research mainly focuses on the one-to-one compatible paradigm, which is limited in learning compatibility among multiple models. We propose a Switchable representation learning Framework with Self-Compatibility (SFSC). SFSC generates a series of compatible sub-models with different capacities through one training process. The optimization of sub-models faces gradients conflict, and we mitigate this problem from the perspective of the magnitude and direction. We adjust the priorities of sub-models dynamically through uncertainty estimation to co-optimize sub-models properly. Besides, the gradients with conflicting directions are projected to avoid mutual interference. SFSC achieves state-of-the-art performance on the evaluated datasets., Comment: Accepted by CVPR 2023

In general, data contain noises which come from faulty instruments, flawed measurements or faulty communication. Learning with data in the context of classification or regression is inevitably affected by noises in the data. In order to remove or greatly reduce the impact of noises, we introduce the ideas of fuzzy membership functions and the Laplacian twin support vector machine (Lap-TSVM). A formulation of the linear intuitionistic fuzzy Laplacian twin support vector machine (IFLap-TSVM) is presented. Moreover, we extend the linear IFLap-TSVM to the nonlinear case by kernel function. The proposed IFLap-TSVM resolves the negative impact of noises and outliers by using fuzzy membership functions and is a more accurate reasonable classifier by using the geometric distribution information of labeled data and unlabeled data based on manifold regularization. Experiments with constructed artificial datasets, several UCI benchmark datasets and MNIST dataset show that the IFLap-TSVM has better classification accuracy than other state-of-the-art twin support vector machine (TSVM), intuitionistic fuzzy twin support vector machine (IFTSVM) and Lap-TSVM., Mathematical Physics, Delft Institute of Applied Mathematics

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2020., This dissertation consists of essays addressing the interactions between sovereign default risk and the household sector. Each chapter employs empirical, theoretical, and quantitative macroeconomic methods. The first chapter studies the impact of income inequality on sovereign spreads under elastic labor and endogenous taxation. I first document that high pretax income inequality is associated with high spreads both across countries and across U.S. states. I then develop a sovereign default model with endogenous progressive taxation and heterogeneous labor in productivity and migration cost. The government chooses the optimal combination of tax and debt, considering their interaction. Progressive taxes redistribute income but discourage labor supply and induce emigration, eroding the tax base and the government’s ability to repay debt. Default risk increases sovereign spreads and borrowing costs. Thus, the government faces a trade-off between redistribution and spreads. In more unequal economies, the government opts for more redistribution and higher spreads. With the model parameterized to state-level data, I find that income inequality is an important determinant of spreads, generating 20 percent higher spreads compared with a model without income inequality. In a recession, more unequal economies suffer a larger increase in spreads. The second chapter studies the salient features of sovereign debt crises: persistent economic declines, spiking spreads, and outflows of capital and workers. We first document a net worker outflow accompanies a rise in sovereign debt spreads. We then develop a sovereign default model with migration choice and capital accumulation. The model features two-way feedback. Default risk lowers workers’ welfare and induces emigration, which in turn intensifies default risk by lowering tax base and investment. Compared with a model without migration, our model generates a higher default risk, lower investment, and a more profound and prolonged

In the originally published version, this manuscript included an error related to indicating the corresponding author within the author list. This has now been corrected online to reflect the fact that author Jun Luo is the corresponding author of the article.

In general, data contain noises which come from faulty instruments, flawed measurements or faulty communication. Learning with data in the context of classification or regression is inevitably affected by noises in the data. In order to remove or greatly reduce the impact of noises, we introduce the ideas of fuzzy membership functions and the Laplacian twin support vector machine (Lap-TSVM). A formulation of the linear intuitionistic fuzzy Laplacian twin support vector machine (IFLap-TSVM) is presented. Moreover, we extend the linear IFLap-TSVM to the nonlinear case by kernel function. The proposed IFLap-TSVM resolves the negative impact of noises and outliers by using fuzzy membership functions and is a more accurate reasonable classifier by using the geometric distribution information of labeled data and unlabeled data based on manifold regularization. Experiments with constructed artificial datasets, several UCI benchmark datasets and MNIST dataset show that the IFLap-TSVM has better classification accuracy than other state-of-the-art twin support vector machine (TSVM), intuitionistic fuzzy twin support vector machine (IFTSVM) and Lap-TSVM., Mathematical Physics

PR, IFPRI3; ISI; CRP4; 2 Promoting Healthy Diets and Nutrition for all; Capacity Strengthening; Advancing Research on Nutrition and Agriculture (ARENA); Feed the Future Innovation Lab for Nutrition; UNFSS, PHND; A4NH, CGIAR Research Programs on Agriculture for Nutrition and Health (A4NH), Many policies and programs aim to bring nutritious diets within reach of the poor. This paper uses retail prices and nutrient composition for 671 foods and beverages to compute the daily cost of essential nutrients required for an active and healthy life in 177 countries around the world. We compare this minimum cost of nutrient adequacy with the subsistence cost of dietary energy and per-capita spending on all goods and services, to identify stylized facts about how diet cost and affordability relate to economic development and nutrition outcomes. On average, the most affordable nutrient adequate diet exceeds the cost of adequate energy by a factor of 2.66, costing US$1.35 per day to meet median requirements of healthy adult women in 2011. Affordability is lowest in Sub-Saharan Africa. The sensitivity of diet costs to each requirement reveals the high cost of staying within acceptable macronutrient ranges, particularly the upper limit for carbohydrates. Among micronutrients, total diet costs are most sensitive to requirements for calcium as well as vitamins A, C, E, B12, folate and riboflavin. On average, about 5% of dietary energy in the least-cost nutrient adequate diets is derived from animal source foods, with small quantities of meat and fish. Over 70% of all animal products in least-cost diets is eggs and dairy, but only in upper-middle and high-income countries. In lower income countries where egg and dairy prices are significantly higher, they are replaced by larger volumes of vegetal foods. When controlling for national income, diet costs are most significantly correlated with rural travel times and rural electrification. These data suggest opportunities for targeted policies and programs that reduce market prices and the cost of nutritious diets, while improving affordability through nutrition assistance, safety nets and higher earnings among low-income households.

Visual retrieval system faces frequent model update and deployment. It is a heavy workload to re-extract features of the whole database every time.Feature compatibility enables the learned new visual features to be directly compared with the old features stored in the database. In this way, when updating the deployed model, we can bypass the inflexible and time-consuming feature re-extraction process. However, the old feature space that needs to be compatible is not ideal and faces the distribution discrepancy problem with the new space caused by different supervision losses. In this work, we propose a global optimization Dual-Tuning method to obtain feature compatibility against different networks and losses. A feature-level prototype loss is proposed to explicitly align two types of embedding features, by transferring global prototype information. Furthermore, we design a component-level mutual structural regularization to implicitly optimize the feature intrinsic structure. Experimental results on million-scale datasets demonstrate that our Dual-Tuning is able to obtain feature compatibility without sacrificing performance. (Our code will be avaliable at https://github.com/yanbai1993/Dual-Tuning)

In object re-identification (ReID), the development of deep learning techniques often involves model updates and deployment. It is unbearable to re-embedding and re-index with the system suspended when deploying new models. Therefore, backward-compatible representation is proposed to enable "new" features to be compared with "old" features directly, which means that the database is active when there are both "new" and "old" features in it. Thus we can scroll-refresh the database or even do nothing on the database to update. The existing backward-compatible methods either require a strong overlap between old and new training data or simply conduct constraints at the instance level. Thus they are difficult in handling complicated cluster structures and are limited in eliminating the impact of outliers in old embeddings, resulting in a risk of damaging the discriminative capability of new features. In this work, we propose a Neighborhood Consensus Contrastive Learning (NCCL) method. With no assumptions about the new training data, we estimate the sub-cluster structures of old embeddings. A new embedding is constrained with multiple old embeddings in both embedding space and discrimination space at the sub-class level. The effect of outliers diminished, as the multiple samples serve as "mean teachers". Besides, we also propose a scheme to filter the old embeddings with low credibility, further improving the compatibility robustness. Our method ensures backward compatibility without impairing the accuracy of the new model. And it can even improve the new model's accuracy in most scenarios., Comment: Accepted by AAAI 2022

Existing public person Re-Identification~(ReID) datasets are small in modern terms because of labeling difficulty. Although unlabeled surveillance video is abundant and relatively easy to obtain, it is unclear how to leverage these footage to learn meaningful ReID representations. In particular, most existing unsupervised and domain adaptation ReID methods utilize only the public datasets in their experiments, with labels removed. In addition, due to small data sizes, these methods usually rely on fine tuning by the unlabeled training data in the testing domain to achieve good performance. Inspired by the recent progress of large-scale self-supervised image classification using contrastive learning, we propose to learn ReID representation from large-scale unlabeled surveillance video alone. Assisted by off-the-shelf pedestrian detection tools, we apply the contrastive loss at both the image and the tracklet levels. Together with a principal component analysis step using camera labels freely available, our evaluation using a large-scale unlabeled dataset shows far superior performance among unsupervised methods that do not use any training data in the testing domain. Furthermore, the accuracy improves with the data size and therefore our method has great potential with even larger and more diversified datasets.

TianQin is a proposed space-based gravitational wave observatory. It is designed to detect the gravitational wave signals in the frequency range of 0.1 mHz -- 1 Hz. At a geocentric distance of $10^5$ km, the plasma in the earth magnetosphere will contribute as the main source of environmental noises. Here, we analyze the acceleration noises that are caused by the magnetic field of space plasma for the test mass of TianQin. The real solar wind data observed by the Advanced Composition Explorer are taken as the input of the magnetohydrodynamic simulation. The Space Weather Modeling Framework is used to simulate the global magnetosphere of the earth, from which we obtain the plasma and magnetic field parameters on the detector's orbits. We calculate the time series of the residual accelerations and the corresponding amplitude spectral densities on these orbit configurations. We find that the residual acceleration produced by the interaction between the TM's magnetic moment induced by the space magnetic field and the spacecraft magnetic field ($\bm{a}_{\rm M1}$) is the dominant term, which can approach $10^{-15}$ m/s$^2$/Hz$^{1/2}$ at $f \approx$ 0.2 mHz for the nominal values of the magnetic susceptibility ($\chi_{\rm m} = 10^{-5}$) and the magnetic shielding factor ($\xi_{\rm m} = 10$) of the test mass. The ratios between the amplitude spectral density of the acceleration noise caused by the space magnetic field and the preliminary goal of the inertial sensor are 0.38 and 0.08 at 1 mHz and 10 mHz, respectively. We discuss the further reduction of this acceleration noise by decreasing $\chi_{\rm m}$ and/or increasing $\xi_{\rm m}$ in the future instrumentation development for TianQin., Comment: 21 pages, 9 figures; revised version, to be published in Classical and Quantum Gravity

Author Posting. © American Geophysical Union, 2020. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Journal of Geophysical Research: Oceans 125(8), (2020): e2020JC016258, doi:10.1029/2020JC016258., This study assessed the effects of typhoons on sea surface pCO2 and CO2 flux in the northern South China Sea (SCS). During the passage of three major typhoons from May to August 2013, sea surface pCO2, surface seawater temperature (SST), and other meteorological parameters were continuously measured on a moored buoy. Surface water in the region was a source of CO2 to the atmosphere with large variations ranging from hours to months. SST was the primary factor controlling the variation of surface pCO2 through most of the time period. Typhoons are seen to impact surface pCO2 in three steps: first by cooling, thus decreasing surface pCO2, and then by causing vertical mixing that brings up deep, high‐CO2 water, and lastly triggering net uptake of CO2 due to the nutrients brought up in this deep water. The typhoons of this study primarily impacted air‐sea CO2 flux via increasing wind speeds. The mean CO2 flux during a typhoon ranged from 3.6 to 5.4 times the pretyphoon mean flux. The magnitude of the CO2 flux during typhoons was strongly inversely correlated with the typhoon center distance. The effect of typhoons accounted for 22% of the total CO2 flux in the study period, during which typhoons occurred only 9% of the time. It was estimated that typhoons enhanced annual CO2 efflux by 23–56% in the northern SCS during the last decade. As such, tropical cyclones may play a large and increasingly important role in controlling CO2 fluxes in a warmer and stormier ocean of the future., This study was supported by the Marine Public Welfare Project of China (Grant 200905012), the Scientific Research Fund of the Second Institute of Oceanography of China (Grant JT1502), the Global Change and Air‐Sea Interaction project of China (Grant GASI‐03‐01‐02‐02), and the National Natural Sciences Foundation of China (Grant 91128212)., 2021-02-03

Author Posting. © American Geophysical Union, 2020. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Journal of Geophysical Research: Oceans 125(8), (2020): e2020JC016258, doi:10.1029/2020JC016258., This study assessed the effects of typhoons on sea surface pCO2 and CO2 flux in the northern South China Sea (SCS). During the passage of three major typhoons from May to August 2013, sea surface pCO2, surface seawater temperature (SST), and other meteorological parameters were continuously measured on a moored buoy. Surface water in the region was a source of CO2 to the atmosphere with large variations ranging from hours to months. SST was the primary factor controlling the variation of surface pCO2 through most of the time period. Typhoons are seen to impact surface pCO2 in three steps: first by cooling, thus decreasing surface pCO2, and then by causing vertical mixing that brings up deep, high‐CO2 water, and lastly triggering net uptake of CO2 due to the nutrients brought up in this deep water. The typhoons of this study primarily impacted air‐sea CO2 flux via increasing wind speeds. The mean CO2 flux during a typhoon ranged from 3.6 to 5.4 times the pretyphoon mean flux. The magnitude of the CO2 flux during typhoons was strongly inversely correlated with the typhoon center distance. The effect of typhoons accounted for 22% of the total CO2 flux in the study period, during which typhoons occurred only 9% of the time. It was estimated that typhoons enhanced annual CO2 efflux by 23–56% in the northern SCS during the last decade. As such, tropical cyclones may play a large and increasingly important role in controlling CO2 fluxes in a warmer and stormier ocean of the future., This study was supported by the Marine Public Welfare Project of China (Grant 200905012), the Scientific Research Fund of the Second Institute of Oceanography of China (Grant JT1502), the Global Change and Air‐Sea Interaction project of China (Grant GASI‐03‐01‐02‐02), and the National Natural Sciences Foundation of China (Grant 91128212)., 2021-02-03

Haiyan Gao,1,2 Yan Bai,1,2 Lijuan Chen,1,2 Georges Ei Fakhri,3 Meiyun Wang1,2 1Henan Provincial People’s Hospital & Zhengzhou University People’s Hospital, Zhengzhou 450003, People’s Republic of China; 2Henan Key Laboratory of Neurological Imaging, Zhengzhou University, Zhengzhou 450003, People’s Republic of China; 3Gordon Center for Medical Imaging, Radiology, Massachusettes General Hospital, Harvard Medical School, Boston, MA 02114, USACorrespondence: Meiyun Wang Email mywang@ha.edu.cnBackground and Purpose: The development of multiple drug resistance (MDR) to chemotherapy and single modal therapy remains unsatisfied for the eradication of tumor, which are major obstacles in cancer therapy. This novel system with excellent characteristics for inhibition of P-glycoprotein (P-gp), and for near-infrared fluorescence (NIRF) imaging-guided chemo-photothermal therapy (PTT), has been identified as a promising way to MDR and achieve synergistic cancer therapy.Methods: In this study, we successfully synthesized a multifunctional theranostic system, which was developed through FDA-approved self-assembling drugs, which contain anticancer drug doxorubicin (Dox), imaging and high photothermal conversion drug indocyanine green (ICG) and P-gp regulator TPGS (the system named T/Dox-ICG). We studied the characterization of T/Dox-ICG NPs, including the TEM, SEM, DLS, UV-vis-NIR, zeta potential, CLSM, in vitro FL imaging, in vitro photothermal effect, in vitro Dox and ICG release. We used CLSM to verify the location of intracellular distribution of Dox in SCG 7901/VCR cells, Western blot was performed to demonstrate the TPGS-mediated inhibition of P-gp. And, the cytotoxicity of materials against SCG 7901/VCR cells was studied by the MTT assay.Results: The TEM showed the T/Dox-ICG NPs had good monodispersity with diameters of 19.03 nm, Dox and ICG could be released constantly from T/Dox-ICG NPs in vitro. In vitro cell experiments demons

Author Posting. © American Geophysical Union, 2020. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Journal of Geophysical Research: Oceans 125(8), (2020): e2020JC016258, doi:10.1029/2020JC016258., This study assessed the effects of typhoons on sea surface pCO2 and CO2 flux in the northern South China Sea (SCS). During the passage of three major typhoons from May to August 2013, sea surface pCO2, surface seawater temperature (SST), and other meteorological parameters were continuously measured on a moored buoy. Surface water in the region was a source of CO2 to the atmosphere with large variations ranging from hours to months. SST was the primary factor controlling the variation of surface pCO2 through most of the time period. Typhoons are seen to impact surface pCO2 in three steps: first by cooling, thus decreasing surface pCO2, and then by causing vertical mixing that brings up deep, high‐CO2 water, and lastly triggering net uptake of CO2 due to the nutrients brought up in this deep water. The typhoons of this study primarily impacted air‐sea CO2 flux via increasing wind speeds. The mean CO2 flux during a typhoon ranged from 3.6 to 5.4 times the pretyphoon mean flux. The magnitude of the CO2 flux during typhoons was strongly inversely correlated with the typhoon center distance. The effect of typhoons accounted for 22% of the total CO2 flux in the study period, during which typhoons occurred only 9% of the time. It was estimated that typhoons enhanced annual CO2 efflux by 23–56% in the northern SCS during the last decade. As such, tropical cyclones may play a large and increasingly important role in controlling CO2 fluxes in a warmer and stormier ocean of the future., This study was supported by the Marine Public Welfare Project of China (Grant 200905012), the Scientific Research Fund of the Second Institute of Oceanography of China (Grant JT1502), the Global Change and Air‐Sea Interaction project of China (Grant GASI‐03‐01‐02‐02), and the National Natural Sciences Foundation of China (Grant 91128212)., 2021-02-03

TianQin is a planned space-based gravitational wave (GW) observatory consisting of three Earth-orbiting satellites with an orbital radius of about $10^5 \, {\rm km}$. The satellites will form an equilateral triangle constellation the plane of which is nearly perpendicular to the ecliptic plane. TianQin aims to detect GWs between $10^{-4} \, {\rm Hz}$ and $1 \, {\rm Hz}$ that can be generated by a wide variety of important astrophysical and cosmological sources, including the inspiral of Galactic ultra-compact binaries, the inspiral of stellar-mass black hole binaries, extreme mass ratio inspirals, the merger of massive black hole binaries, and possibly the energetic processes in the very early universe and exotic sources such as cosmic strings. In order to start science operations around 2035, a roadmap called the 0123 plan is being used to bring the key technologies of TianQin to maturity, supported by the construction of a series of research facilities on the ground. Two major projects of the 0123 plan are being carried out. In this process, the team has created a new-generation $17 \, {\rm cm}$ single-body hollow corner-cube retro-reflector which was launched with the QueQiao satellite on 21 May 2018; a new laser-ranging station equipped with a $1.2 \, {\rm m}$ telescope has been constructed and the station has successfully ranged to all five retro-reflectors on the Moon; and the TianQin-1 experimental satellite was launched on 20 December 2019 - the first-round result shows that the satellite has exceeded all of its mission requirements.

Haiyan Gao,1,2 Yan Bai,1,2 Lijuan Chen,1,2 Georges Ei Fakhri,3 Meiyun Wang1,2 1Henan Provincial People’s Hospital & Zhengzhou University People’s Hospital, Zhengzhou 450003, People’s Republic of China; 2Henan Key Laboratory of Neurological Imaging, Zhengzhou University, Zhengzhou 450003, People’s Republic of China; 3Gordon Center for Medical Imaging, Radiology, Massachusettes General Hospital, Harvard Medical School, Boston, MA 02114, USACorrespondence: Meiyun Wang Email mywang@ha.edu.cnBackground and Purpose: The development of multiple drug resistance (MDR) to chemotherapy and single modal therapy remains unsatisfied for the eradication of tumor, which are major obstacles in cancer therapy. This novel system with excellent characteristics for inhibition of P-glycoprotein (P-gp), and for near-infrared fluorescence (NIRF) imaging-guided chemo-photothermal therapy (PTT), has been identified as a promising way to MDR and achieve synergistic cancer therapy.Methods: In this study, we successfully synthesized a multifunctional theranostic system, which was developed through FDA-approved self-assembling drugs, which contain anticancer drug doxorubicin (Dox), imaging and high photothermal conversion drug indocyanine green (ICG) and P-gp regulator TPGS (the system named T/Dox-ICG). We studied the characterization of T/Dox-ICG NPs, including the TEM, SEM, DLS, UV-vis-NIR, zeta potential, CLSM, in vitro FL imaging, in vitro photothermal effect, in vitro Dox and ICG release. We used CLSM to verify the location of intracellular distribution of Dox in SCG 7901/VCR cells, Western blot was performed to demonstrate the TPGS-mediated inhibition of P-gp. And, the cytotoxicity of materials against SCG 7901/VCR cells was studied by the MTT assay.Results: The TEM showed the T/Dox-ICG NPs had good monodispersity with diameters of 19.03 nm, Dox and ICG could be released constantly from T/Dox-ICG NPs in vitro. In vitro cell experiments demons

TianQin is a planned space-based gravitational wave (GW) observatory consisting of three earth orbiting satellites with an orbital radius of about $10^5~{\rm km}$. The satellites will form a equilateral triangle constellation the plane of which is nearly perpendicular to the ecliptic plane. TianQin aims to detect GWs between $10^{-4}~{\rm Hz}$ and $1~{\rm Hz}$ that can be generated by a wide variety of important astrophysical and cosmological sources, including the inspiral of Galactic ultra-compact binaries, the inspiral of stellar-mass black hole binaries, extreme mass ratio inspirals, the merger of massive black hole binaries, and possibly the energetic processes in the very early universe or exotic sources such as cosmic strings. In order to start science operations around 2035, a roadmap called the 0123 plan is being used to bring the key technologies of TianQin to maturity, supported by the construction of a series of research facilities on the ground. Two major projects of the 0123 plan are being carried out. In this process, the team has created a new generation $17~{\rm cm}$ single-body hollow corner-cube retro-reflector which has been launched with the QueQiao satellite on 21 May 2018; a new laser ranging station equipped with a $1.2~{\rm m}$ telescope has been constructed and the station has successfully ranged to all the five retro-reflectors on the Moon; and the TianQin-1 experimental satellite has been launched on 20 December 2019 and the first round result shows that the satellite has exceeded all of its mission requirements., Comment: 16 pages, 1 figure, invited article to a special section in PTEP

TianQin is a proposed space-based gravitational wave observatory. It is designed to detect the gravitational wave signals in the frequency range of 0.1 mHz -- 1 Hz. At a geocentric distance of $10^5$ km, the plasma in the earth magnetosphere will contribute as the main source of environmental noises. Here, we analyze the acceleration noises that are caused by the magnetic field of space plasma for the test mass of TianQin. The real solar wind data observed by the Advanced Composition Explorer are taken as the input of the magnetohydrodynamic simulation. The Space Weather Modeling Framework is used to simulate the global magnetosphere of the earth, from which we obtain the plasma and magnetic field parameters on the detector's orbits. We calculate the time series of the residual accelerations and the corresponding amplitude spectral densities on these orbit configurations. We find that the residual acceleration produced by the interaction between the TM's magnetic moment induced by the space magnetic field and the spacecraft magnetic field ($\bm{a}_{\rm M1}$) is the dominant term, which can approach $10^{-15}$ m/s$^2$/Hz$^{1/2}$ at $f \approx$ 0.2 mHz for the nominal values of the magnetic susceptibility ($\chi_{\rm m} = 10^{-5}$) and the magnetic shielding factor ($\xi_{\rm m} = 10$) of the test mass. The ratios between the amplitude spectral density of the acceleration noise caused by the space magnetic field and the preliminary goal of the inertial sensor are 0.38 and 0.08 at 1 mHz and 10 mHz, respectively. We discuss the further reduction of this acceleration noise by decreasing $\chi_{\rm m}$ and/or increasing $\xi_{\rm m}$ in the future instrumentation development for TianQin., Comment: 21 pages, 9 figures; revised version, to be published in Classical and Quantum Gravity

Knowledge distillation(KD) aims to improve the performance of a student network by mimicing the knowledge from a powerful teacher network. Existing methods focus on studying what knowledge should be transferred and treat all samples equally during training. This paper introduces the adaptive sample weighting to KD. We discover that previous effective hard mining methods are not appropriate for distillation. Furthermore, we propose Prime-Aware Adaptive Distillation (PAD) by the incorporation of uncertainty learning. PAD perceives the prime samples in distillation and then emphasizes their effect adaptively. PAD is fundamentally different from and would refine existing methods with the innovative view of unequal training. For this reason, PAD is versatile and has been applied in various tasks including classification, metric learning, and object detection. With ten teacher-student combinations on six datasets, PAD promotes the performance of existing distillation methods and outperforms recent state-of-the-art methods., Comment: Accepted by ECCV 2020

Data augmentation in feature space is effective to increase data diversity. Previous methods assume that different classes have the same covariance in their feature distributions. Thus, feature transform between different classes is performed via translation. However, this approach is no longer valid for recent deep metric learning scenarios, where feature normalization is widely adopted and all features lie on a hypersphere. This work proposes a novel spherical feature transform approach. It relaxes the assumption of identical covariance between classes to an assumption of similar covariances of different classes on a hypersphere. Consequently, the feature transform is performed by a rotation that respects the spherical data distributions. We provide a simple and effective training method, and in depth analysis on the relation between the two different transforms. Comprehensive experiments on various deep metric learning benchmarks and different baselines verify that our method achieves consistent performance improvement and state-of-the-art results., Comment: ECCV2020

Unsupervised domain adaptive (UDA) person re-identification (re-ID) is a challenging task due to the missing of labels for the target domain data. To handle this problem, some recent works adopt clustering algorithms to off-line generate pseudo labels, which can then be used as the supervision signal for on-line feature learning in the target domain. However, the off-line generated labels often contain lots of noise that significantly hinders the discriminability of the on-line learned features, and thus limits the final UDA re-ID performance. To this end, we propose a novel approach, called Dual-Refinement, that jointly refines pseudo labels at the off-line clustering phase and features at the on-line training phase, to alternatively boost the label purity and feature discriminability in the target domain for more reliable re-ID. Specifically, at the off-line phase, a new hierarchical clustering scheme is proposed, which selects representative prototypes for every coarse cluster. Thus, labels can be effectively refined by using the inherent hierarchical information of person images. Besides, at the on-line phase, we propose an instant memory spread-out (IM-spread-out) regularization, that takes advantage of the proposed instant memory bank to store sample features of the entire dataset and enable spread-out feature learning over the entire training data instantly. Our Dual-Refinement method reduces the influence of noisy labels and refines the learned features within the alternative training process. Experiments demonstrate that our method outperforms the state-of-the-art methods by a large margin., Comment: 14 pages, 5 figures

Recently proposed L2-norm linear discriminant analysis criterion via the Bhattacharyya error bound estimation (L2BLDA) is an effective improvement of linear discriminant analysis (LDA) for feature extraction. However, L2BLDA is only proposed to cope with vector input samples. When facing with two-dimensional (2D) inputs, such as images, it will lose some useful information, since it does not consider intrinsic structure of images. In this paper, we extend L2BLDA to a two-dimensional Bhattacharyya bound linear discriminant analysis (2DBLDA). 2DBLDA maximizes the matrix-based between-class distance which is measured by the weighted pairwise distances of class means and meanwhile minimizes the matrix-based within-class distance. The weighting constant between the between-class and within-class terms is determined by the involved data that makes the proposed 2DBLDA adaptive. In addition, the criterion of 2DBLDA is equivalent to optimizing an upper bound of the Bhattacharyya error. The construction of 2DBLDA makes it avoid the small sample size problem while also possess robustness, and can be solved through a simple standard eigenvalue decomposition problem. The experimental results on image recognition and face image reconstruction demonstrate the effectiveness of the proposed methods.

PR, IFPRI3; ISI; CRP4; Advancing Research on Nutrition and Agriculture (ARENA); 2 Promoting Healthy Diets and Nutrition for all; UNFSS, DSGD; PHND; A4NH, CGIAR Research Program on Agriculture for Nutrition and Health (A4NH), The EAT–Lancet Commission drew on all available nutritional and environmental evidence to construct the first global benchmark diet capable of sustaining health and protecting the planet, but it did not assess dietary affordability. We used food price and household income data to estimate affordability of EAT–Lancet benchmark diets, as a first step to guiding interventions to improve diets around the world.

Non-PR, IFPRI5; CRP4; Advancing Research on Nutrition and Agriculture (ARENA); 1 Fostering Climate-Resilient and Sustainable Food Supply; 2 Promoting Healthy Diets and Nutrition for all; 3 Building Inclusive and Efficient Markets, Trade Systems, and Food Industry, DSGD; PHND; A4NH, CGIAR Research Program on Agriculture for Nutrition and Health (A4NH), Messages around healthy eating are nothing new – but nowadays we are increasingly told to eat well not just for our own health but also for the planet. This debate was given a boost at the beginning of 2019 with the publication of the EAT-Lancet Commission on the world's diet, which called for a “great food transformation”. It condemned the world’s "faulty food system" which sees nearly one billion people go hungry, almost two billion eat too much of the wrong type of food, and which puts unsustainable pressure on the planet.

Non-PR, IFPRI5; Advancing Research on Nutrition and Agriculture (ARENA); CRP4, PHND; A4NH, 54 pages, CGIAR Research Program on Agriculture for Nutrition and Health (A4NH), Food prices vary widely around the world, depending on the local cost of retail services and supply chains as well as farm supply and the border prices of tradable commodities. This study measures the cost of the most affordable nutritionally adequate diet in each country, relative to the cost of dietary energy and other benchmarks such as national income and poverty lines, so as to identify development paths associated with lower cost access to the nutrients needed for a healthy and active life. We use prices for 195 standardized food and beverage items across 164 countries in 2011 collected by the World Bank's International Comparison Project (ICP), matched with data on these items' composition in terms of 21 essential nutrients and each nutrient's lower and upper limits for a healthy adult woman. Using a subsample of 134 countries for which economic structure data are available, we find that the cost of nutrient adequacy is highest in poorer and middle-income countries, and is highe

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2016. This work was supported by a dissertation committee consisting of Professor Yan Bai (co-advisor) and Professor Toni Whited (co-advisor), and Professor Olga Itenberg of the Department of Finance at the Simon Business School. Chapter 2 was written in collaboration with Yan Bai and Dan Lu of the Department of Economics. All other work conducted for the dissertation was completed by the student independently., This dissertation studies the macroeconomic consequences of financial frictions via their roles in determining the capital structures of firms and financial institutions. It consists of two papers in this particular field. The first paper focuses on the capital structure decisions of financial intermediaries and their macroeconomic implications. In this paper, titled "Uncertainty and the Shadow Banking Crisis: A Structural Estimation", I examine the impact of asset return uncertainty on the financing and leverage decisions of shadow banks. Shadow banks play an important role in the modern financial system and are arguably the source of key vulnerabilities leading to the 2007-2009 financial crisis. In this paper, I develop a quantitative framework with endogenous bank default and aggregate uncertainty fluctuation to study the dynamics of shadow banking. I argue that the increase in asset return uncertainty during the crisis results in the spread spike, making it more costly for shadow banks to roll over their debt in the short-term debt market. As a result, these banks are forced to deleverage, leading to a decrease in the credit supply. The model is estimated using a bank-level dataset of shadow banks in the United States. The findings show that uncertainty shocks are able to generate statistics and pathways of leverage, spread, and assets which closely match those observed in the data. Maturity mismatch and asset firesales amplify the impact of the uncertainty shocks. First moment shocks alone can not reproduce the large interbank spread spike, dramatic deleveraging and contraction of the US shadow banking sector during the crisis. The model also allows for policy experiments. I analyze how unconventional monetary policies can help to counter the rise in the interbank spread, thus stabilizing the credit supply. Taking into consideration of bank moral hazard, I find that government bailout might be counterproductive as it might result in more aggressive risk-taking

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2016., This dissertation studies sovereign credit ratings both empirically and theoretically. The first chapter uses an event study approach to examine the impact of sovereign credit ratings' announcements on interest rate spreads. I find that only rating downgrades are associated with statistically significant changes in the spread at the time an announcement is made. Moreover, on the days leading up to an announcement, there are statistically significant changes in the spread for both downgrades and upgrades, which could be either due to market participants anticipating the rating change or other events affecting the spread. Finally, this chapter also studies how changes in the spread differ across announcements of different type. The second chapter begins by documenting two empirical regularities of sovereign credit ratings: on average, a country's spread increases sharply as its rating worsens and spread dispersion within a rating level is larger for worse ratings. Moreover, it shows that countries that suffer a downgrade face a higher probability of a sudden stop in capital in flows in the near future, even after accounting for fundamentals. I then incorporate sovereign credit ratings into a sovereign default model. Ratings play two roles in the model: first, they provide information about the country's fundamentals, and second, a downgrade increases the probability of a sudden stop. In a quantitative exercise calibrated to Greece, this model can capture the empirical regularities of ratings documented previously. Inspired by recent regulations in the European Union, this chapter studies the consequences of restricting the publication of ratings using the model. Under the benchmark calibration and without including their potential effects on the overall probability of sudden stops, these regulations increase the welfare of a country because they (i) provide insurance in the form of cheaper borrowing during times of bad economic fundamentals, and (ii) at the mar- gin

This paper presents a rigorous validation of five widely used global land cover products, i.e., GLCC (Global Land Cover Characterization), UMd (University of Maryland land cover product), GLC2000 (Global Land Cover 2000 project data), MODIS LC (Moderate Resolution Imaging Spectro-radiometer Land Cover product) and GlobCover (GLOBCOVER land cover product), and a national land cover map GLCD-2005 (Geodata Land Cover Dataset for year 2005) against an independent reference data set over China. The land cover reference data sets in three epochs (1990, 2000, and 2005) were collected on a web-based prototype system using a sampling-based labeling approach. Results show that, in China, the highest overall accuracy is observed in GLCD-2005 (72.3%), followed by MODIS LC (68.9%), GLC2000 (65.2%), GlobCover (57.7%) and GLCC (57.2%), while UMd has the lowest accuracy (48.6%); all of the products performed best in representing “Trees” and “Others”, well with “Grassland” and “Cropland”, but problematic with “Water” and “Urban” across China in general. Moreover, in respect of GLCD-2005, there are significant accuracy differences across seven geographical locations of China, ranging from 46.3% in the Southwest, 77.5% in the South, 79.2% in the Northwest, 80.8% in the North, 81.8% in the Northeast, 82.6% in the Central, to 89.0% in the East. This study indicates that a regionally focused land cover map would in fact be more accurate than extracting the same region from a globally produced map.

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2015., This thesis examines the role of housing as collateral and occupational choices in aggregate economies. Chapter 1 studies the impact of house price fuctuations on small businesses. The unprecedented upheavals in the U.S. housing market since 2000 and corresponding oscillations in home equity values profoundly affected the net worth and borrowing capacities of individual households. I develop a quantitative model where changes in house prices influence households' borrowing capacities, which in turn influence the entry-exit and expansion-contraction decisions of small business owners. I show that the housing collateral effect can explain the empirically observed strong correlation between house prices and small business activities (as measured by the number of businesses and the number of employees in the small business sector). Next, I conduct an experiment to measure how much of the shrinkage in small business activities during the recent recession can be explained by the housing collateral effect. I argue that the decrease in the value of housing as collateral, following the housing market crash in 2007, can account for 53 percent of the decrease in the number of small businesses and 98 percent of the decrease in the level of small business employment. In Chapter 2, I present an occupational choice model which emphasizes the use of housing as collateral and apply the model to examine the magnitude of the effect of a housing boom on economic growth in countries at different stages of financial development. The model results are twofold. First, a housing boom mitigates capital misallocation which results from an incomplete financial system, by expanding a business owner's borrowing capacity through an increase in collateral value, and thus boosts economy. Second, the impact of a housing boom is greater in countries with less developed financial systems. I provide empirical evidence to support the model results. To get around an endogeneity issue regarding housing bo

Yongjiu Lv,* Jingjing Li,* Huali Chen, Yan Bai, Liangke Zhang Chongqing Key Laboratory of Biochemistry and Molecular Pharmacology, Chongqing Research Center for Pharmaceutical Engineering, College of Pharmacy, Chongqing Medical University, Chongqing, China *These authors contributed equally to this work Abstract: In this study, a glycyrrhetinic acid-functionalized mesoporous silica nanoparticle (MSN-GA) was prepared for active tumor targeting. MSN-GA exhibited satisfactory loading capacity for insoluble drugs, uniform size distribution, and specific tumor cell targeting. Glycyrrhetinic acid, a hepatocellular carcinoma-targeting group, was covalently decorated on the surface of MSN via an amido bond. The successful synthesis of MSN-GA was validated by the results of Fourier transform infrared spectroscopy, transmission electron microscopy (TEM), and zeta potential measurement. TEM images revealed the spherical morphology and uniform size distribution of the naked MSN and MSN-GA. Curcumin (CUR), an insoluble model drug, was loaded into MSN-GA (denoted as MSN-GA-CUR) with a high-loading capacity (8.78%±1.24%). The results of the in vitro cellular experiment demonstrated that MSN-GA-CUR significantly enhanced cytotoxicity and cellular uptake toward hepatocellular carcinoma (HepG2) cells via a specific GA receptor-mediated endocytosis mechanism. The results of this study provide a promising nanoplatform for the targeting of hepatocellular carcinoma. Keywords: active tumor targeting, silica nanoparticle, drug loading capacity, HepG2 cells

Yongjiu Lv,* Jingjing Li,* Huali Chen, Yan Bai, Liangke Zhang Chongqing Key Laboratory of Biochemistry and Molecular Pharmacology, Chongqing Research Center for Pharmaceutical Engineering, College of Pharmacy, Chongqing Medical University, Chongqing, China *These authors contributed equally to this work Abstract: In this study, a glycyrrhetinic acid-functionalized mesoporous silica nanoparticle (MSN-GA) was prepared for active tumor targeting. MSN-GA exhibited satisfactory loading capacity for insoluble drugs, uniform size distribution, and specific tumor cell targeting. Glycyrrhetinic acid, a hepatocellular carcinoma-targeting group, was covalently decorated on the surface of MSN via an amido bond. The successful synthesis of MSN-GA was validated by the results of Fourier transform infrared spectroscopy, transmission electron microscopy (TEM), and zeta potential measurement. TEM images revealed the spherical morphology and uniform size distribution of the naked MSN and MSN-GA. Curcumin (CUR), an insoluble model drug, was loaded into MSN-GA (denoted as MSN-GA-CUR) with a high-loading capacity (8.78%±1.24%). The results of the in vitro cellular experiment demonstrated that MSN-GA-CUR significantly enhanced cytotoxicity and cellular uptake toward hepatocellular carcinoma (HepG2) cells via a specific GA receptor-mediated endocytosis mechanism. The results of this study provide a promising nanoplatform for the targeting of hepatocellular carcinoma. Keywords: active tumor targeting, silica nanoparticle, drug loading capacity, HepG2 cells

This paper provides an overview of the on-going compact descriptors for video analysis standard (CDVA) from the ISO/IEC moving pictures experts group (MPEG). MPEG-CDVA targets at defining a standardized bitstream syntax to enable interoperability in the context of video analysis applications. During the developments of MPEGCDVA, a series of techniques aiming to reduce the descriptor size and improve the video representation ability have been proposed. This article describes the new standard that is being developed and reports the performance of these key technical contributions., Comment: 4 figures, 4 tables

Fine-grained visual recognition aims to capture discriminative characteristics amongst visually similar categories. The state-of-the-art research work has significantly improved the fine-grained recognition performance by deep metric learning using triplet network. However, the impact of intra-category variance on the performance of recognition and robust feature representation has not been well studied. In this paper, we propose to leverage intra-class variance in metric learning of triplet network to improve the performance of fine-grained recognition. Through partitioning training images within each category into a few groups, we form the triplet samples across different categories as well as different groups, which is called Group Sensitive TRiplet Sampling (GS-TRS). Accordingly, the triplet loss function is strengthened by incorporating intra-class variance with GS-TRS, which may contribute to the optimization objective of triplet network. Extensive experiments over benchmark datasets CompCar and VehicleID show that the proposed GS-TRS has significantly outperformed state-of-the-art approaches in both classification and retrieval tasks., Comment: 6 pages, 5 figures

Object detection aims to identify instances of semantic objects of a certain class in images or videos. The success of state-of-the-art approaches is attributed to the significant progress of object proposal and convolutional neural networks (CNNs). Most promising detectors involve multi-task learning with an optimization objective of softmax loss and regression loss. The first is for multi-class categorization, while the latter is for improving localization accuracy. However, few of them attempt to further investigate the hardness of distinguishing different sorts of distracting background regions (i.e., negatives) from true object regions (i.e., positives). To improve the performance of classifying positive object regions vs. a variety of negative background regions, we propose to incorporate triplet embedding into learning objective. The triplet units are formed by assigning each negative region to a meaningful object class and establishing class- specific negatives, followed by triplets construction. Over the benchmark PASCAL VOC 2007, the proposed triplet em- bedding has improved the performance of well-known FastRCNN model with a mAP gain of 2.1%. In particular, the state-of-the-art approach OHEM can benefit from the triplet embedding and has achieved a mAP improvement of 1.2%., Comment: 6 pages, 5 figures

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2015., This dissertation contributes to literature of International Economics and, in particular, of Sovereign Default with the study of three distinct issues. In the first chapter, I study the role of international reserves in sovereign debt restructuring episodes when fiscal adjustment is costly. This study departs from the observation that highly indebted developing economies commonly also hold large external reserves. This behavior seems puzzling given that governments in these countries borrow with an interest rate penalty to compensate lenders for default risk. Reducing debt to the same extent as reserves would maintain net liabilities constant while decreasing interest payments. However, holding reserves can have insurance benefits in a financial crisis. To rationalize the levels of international reserves and external debt observed in the data, a standard dynamic model of equilibrium default is extended to include distortionary taxation and debt restructuring. This chapter shows that fiscal adjustments induced by sovereign default can generate large demand for reserves if taxation is distortionary. At the same time, a non-negligible position in reserves modifies the debt restructuring negotiations upon default. A calibrated version of the model produces recovery rate schedules that are increasing with reserves, as seen in the data, being also able to replicate large positions of reserves and debt to GDP. Finally, I study how both mechanisms play a key quantitative role to generate such result, in fact, not including them, produces a counterfactual demand for reserves that is close to zero. In the second chapter, the relationship between labor market distortions and sovereign debt crises is explored. It is noted that risk of sovereign debt default has frequently affected both emerging market and developed economies. Such financial crises are often accompanied with severe declines of employment that are hard to justify using a standard dynamic stochastic model. In thi

Yan-Rong Wang,1 Yi-Ning Qiu,2 Yan Bai,2 Xian-Feng Wang1 1Department of Pediatrics, The Affiliated Shenzhen Third Hospital, Guangdong Medical College, Shenzhen, 2Department of Pediatrics, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, People’s Republic of China Aim: The aim of this study was to investigate the etiological factors, clinical features, and prognostic factors in children with hemophagocytic lymphohistiocytosis (HLH).Methods: Fifty-six children with HLH in Wuhan Union Hospital, People’s Republic of China, were retrospectively analyzed in recent years. We reviewed the medical records of 56 HLH children hospitalized from 2000 to 2013 to identify the possible prognostic factors.Results: In more than half of the cases (64.29%), the etiological factor was found to be infection. Clinical characteristics such as prolonged fever (100.00%), hepatosplenomegaly (95.24%), and pancytopenia (100.00%) were observed. Characteristic laboratory values presented with increased ferritin (64.29%), triglycerides (78.57%), transaminases (80.95%), bilirubin (67.54%), lactate dehydrogenase (95.23%), and decreased fibrinogen (61.90%), natrium (40.48%), and potassium (30.94%). Bone marrow aspiration showed hemophagocytosis in 48 cases (85.71%). Forty-two patients were treated according to HLH-2004 protocol. Out of the 42 patients, 19 cases acquired remission and 13 cases died. Ten cases lost follow-up. High lactate dehydrogenase (>2,000 U/L), high bilirubin (>2 mg/mL), and younger age (<2 years) at the time of diagnosis were adverse prognostic factors. Conclusion: HLH is a life-threatening syndrome caused by complicated etiology. The level of lactate dehydrogenase and bilirubin accompanied with younger age were adverse factors. Keywords: HLH, clinical features, children, prognostic factors