Xiamen University

About: Xiamen University is a education organization based out in Amoy, Fujian, China. It is known for research contribution in the topics: Catalysis & Population. The organization has 50472 authors who have published 54480 publications receiving 1058239 citations. The organization is also known as: Amoy University & Xiàmén Dàxué.

Invariance Matters: Exemplar Memory for Domain Adaptive Person Re-Identification

Abstract: This paper considers the domain adaptive person re-identification (re-ID) problem: learning a re-ID model from a labeled source domain and an unlabeled target domain. Conventional methods are mainly to reduce feature distribution gap between the source and target domains. However, these studies largely neglect the intra-domain variations in the target domain, which contain critical factors influencing the testing performance on the target domain. In this work, we comprehensively investigate into the intra-domain variations of the target domain and propose to generalize the re-ID model w.r.t three types of the underlying invariance, i.e., exemplar-invariance, camera-invariance and neighborhood-invariance. To achieve this goal, an exemplar memory is introduced to store features of the target domain and accommodate the three invariance properties. The memory allows us to enforce the invariance constraints over global training batch without significantly increasing computation cost. Experiment demonstrates that the three invariance properties and the proposed memory are indispensable towards an effective domain adaptation system. Results on three re-ID domains show that our domain adaptation accuracy outperforms the state of the art by a large margin. Code is available at: https://github.com/zhunzhong07/ECN

Predicting Diabetes Mellitus With Machine Learning Techniques.

Abstract: Diabetes mellitus is a chronic disease characterized by hyperglycemia. It may cause many complications. According to the growing morbidity in recent years, in 2040, the world's diabetic patients will reach 642 million, which means that one of the ten adults in the future is suffering from diabetes. There is no doubt that this alarming figure needs great attention. With the rapid development of machine learning, machine learning has been applied to many aspects of medical health. In this study, we used decision tree, random forest and neural network to predict diabetes mellitus. The dataset is the hospital physical examination data in Luzhou, China. It contains 14 attributes. In this study, five-fold cross validation was used to examine the models. In order to verity the universal applicability of the methods, we chose some methods that have the better performance to conduct independent test experiments. We randomly selected 68994 healthy people and diabetic patients' data, respectively as training set. Due to the data unbalance, we randomly extracted 5 times data. And the result is the average of these five experiments. In this study, we used principal component analysis (PCA) and minimum redundancy maximum relevance (mRMR) to reduce the dimensionality. The results showed that prediction with random forest could reach the highest accuracy (ACC = 0.8084) when all the attributes were used.

Understanding COVID-19 vaccine demand and hesitancy: A nationwide online survey in China.

Abstract: Background This study attempts to understand coronavirus disease 2019 (COVID-19) vaccine demand and hesitancy by assessing the public’s vaccination intention and willingness-to-pay (WTP). Confidence in COVID-19 vaccines produced in China and preference for domestically-made or foreign-made vaccines was also investigated. Methods A nationwide cross-sectional, self-administered online survey was conducted on 1–19 May 2020. The health belief model (HBM) was used as a theoretical framework for understanding COVID-19 vaccination intent and WTP. Results A total of 3,541 complete responses were received. The majority reported a probably yes intent (54.6%), followed by a definite yes intent (28.7%). The perception that vaccination decreases the chances of getting COVID-19 under the perceived benefit construct (OR = 3.14, 95% CI 2.05–4.83) and not being concerned about the efficacy of new COVID-19 vaccines under the perceived barriers construct (OR = 1.65, 95% CI 1.31–2.09) were found to have the highest significant odds of a definite intention to take the COVID-19 vaccine. The median (interquartile range [IQR]) of WTP for COVID-19 vaccine was CNY¥200/US$28 (IQR CNY¥100–500/USD$14–72). The highest marginal WTP for the vaccine was influenced by socio-economic factors. The majority were confident (48.7%) and completely confident (46.1%) in domestically-made COVID-19 vaccine. 64.2% reported a preference for a domestically-made over foreign-made COVID-19 vaccine. Conclusions The findings demonstrate the utility of HBM constructs in understanding COVID-19 vaccination intent and WTP. It is important to improve health promotion and reduce the barriers to COVID-19 vaccination.

Amine-Assisted Synthesis of Concave Polyhedral Platinum Nanocrystals Having {411} High-Index Facets

Abstract: High-index surfaces of a face-centered cubic metal (e.g., Pd, Pt) have a high density of low-coordinated surface atoms and therefore possess enhanced catalysis activity in comparison with low-index faces. However, because of their high surface energy, the challenge of chemically preparing metal nanocrystals having high-index facets remains. We demonstrate in this work that introducing amines as the surface controller allows concave Pt nanocrystals having {411} high-index facets to be prepared through a facile wet-chemical route. The as-prepared Pt nanocrystals display a unique octapod morphology with {411} facets. The presence of high-index {411} exposed facets endows the concave Pt nanocrystals with excellent electrocatalytic activity in the oxidation of both formic acid and ethanol.

p-n Heterojunction photoelectrodes composed of Cu2O-loaded TiO2 nanotube arrays with enhanced photoelectrochemical and photoelectrocatalytic activities

Abstract: Cu2O/TiO2 p–n heterojunction photoelectrodes were prepared by depositing different amounts of p-type Cu2O nanoparticles on n-type TiO2 nanotube arrays (i.e., forming Cu2O/TiO2 composite nanotubes) via an ultrasonication-assisted sequential chemical bath deposition. The success of deposition of Cu2O nanoparticles was corroborated by structural and composition characterizations. The enhanced absorption in the visible light region was observed in Cu2O/TiO2 composite nanotubes. The largely improved separation of photogenerated electrons and holes was revealed by photocurrent measurements. Consequently, Cu2O/TiO2 heterojunction photoelectrodes exhibited a more effective photoconversion capability than TiO2 nanotubes alone in photoelectrochemical measurements. Furthermore, Cu2O/TiO2 composite photoelectrodes also possessed superior photoelectrocatalytic activity and stability in the degradation of Rhodamine B. Intriguingly, by selecting an appropriate bias potential, a synergistic effect between electricity and visible light irradiation can be achieved.