作者机构:
[Wen, Yimin; Xu, Zhi] Guilin Univ Elect Technol, Guangxi Key Lab Image & Graph Intelligent Proc, Guilin 541004, Peoples R China.;[Wen, Yimin; Yi, Haiyang] Guilin Univ Elect Technol, Guangxi Key Lab Trusted Software, Guilin 541004, Peoples R China.;[Li, Yujian] Guilin Univ Elect Technol, Sch Artificial Intelligence, Guilin 541004, Peoples R China.;[Fan, Zhigang] Zebra Technol China Corp, Shanghai 200122, Peoples R China.;[Xue, Yun] Hunan City Univ, Sch Municipal & Surveying Engn, Yiyang 413000, Peoples R China.
通讯机构:
[Yimin Wen] G;Guangxi Key Laboratory of Image and Graphic Intelligent Processing, Guilin University of Electronic Technology, Guilin 541004, China<&wdkj&>Guangxi Key Laboratory of Trusted Software, Guilin University of Electronic Technology, Guilin 541004, China
关键词:
Domain adaptation;Discriminative analysis;Single sample face recognition;Transfer learning;Gallery-sensitive
摘要:
Taking advantage of labeled auxiliary training data whose distribution is similar to the distribution of the gallery, single sample face recognition (SSFR) has achieved encouraging performance. However, in many real-world applications, it is difficult to collect such an auxiliary training dataset, while it may be easier to collect an unlabeled target training dataset whose distribution is similar to the distribution of the gallery and a labeled source training dataset whose distribution may be different to the distribution of the gal-lery. How can these three datasets be effectively leveraged to handle SSFR? To address this issue, this paper proposes a new method of Gallery-Sensitive Single Sample Face Recognition based on Domain Adaptation (GS-DA). First, GS-DA employs the method of TSD (targetize the source domain) to construct a common subspace and a targetized source domain. Secondly, it projects each gallery image into the common subspace and obtains the sparse representation of each gallery image in the common subspace. Thirdly, it reconstructs each gallery image from the targetized source domain to estimate the within-class scatter matrix and the between-class scatter matrix of the gallery. Lastly, it learns a discriminant model by maximizing the sum of the traces of the between-class scatter matrix of the gallery and the between-class scatter matrix of the targetized source domain as well as minimizing the sum of the traces of the total scatter matrix of the gallery and the total scatter matrix of the target training data. The experimental results on five datasets illustrate the superiority of GS-DA in leveraging these three datasets for SSFR. (c) 2020 Elsevier B.V. All rights reserved.
作者机构:
[Ji, Weizhen; Liu, Deer] School of Architectural and Surveying and Mapping Engineering, Jiangxi University of Science and Technology, Ganzhou;341000, China;[Meng, Yifei] School of Geography and Information Engineering, China University of Geosciences (Wuhan), Wuhan;430074, China;[Xue, Yun] School of Municipal and Surveying Engineering, Hunan City University, Yiyang
通讯机构:
[Deer Liu] S;School of Architectural and Surveying and Mapping Engineering, Jiangxi University of Science and Technology, Ganzhou, China
通讯机构:
[Chen, Yaoning; Li, Yuanping] H;Hunan Univ, Coll Environm Sci & Engn, Changsha 410082, Peoples R China.;Hunan Univ, Minist Educ, Key Lab Environm Biol & Pollut Control, Changsha 410082, Peoples R China.;Hunan City Univ, Coll Municipal & Mapping Engn, Yiyang 413000, Hunan, Peoples R China.
摘要:
Visible-light-driven photocatalysis, as an emerging environmental restoration technology, has become a focus of attention recently. In this work, a novel SrCO3-SrTiO3/Ag3PO4 heterojunction was successfully synthesized, and its structure, composition, and optical performance were systematically characterized. In addition, SrTiO3/Ag3PO4 was also synthesized, and its type II heterojunction had certain defects. The active species trapping experiments of SrCO3-SrTiO3/Ag3PO4 showed that center dot OH and center dot O-2(-) participated during the reaction. The photodegradation efficiency of SrCO3-SrTiO3/Ag3PO4 heterojunction was 1.4 times and 2.0 times that of Ag3PO4 and SrCO3-SrTiO3, respectively. A novel Z-scheme mechanism could be used to explain the photoreaction process of SrCO3-SrTiO3/Ag3PO4. The excellent photocatalytic property was due to the enhancement of visible-light absorption, the effective separation of photogenerated charges, and the generation of abundant active species.
摘要:
The triple-bed circulating fluidized bed gasifier is a new type of the gasification process in which the combustion process, pyrolysis process, and gasification process of the fuel are carried out in different reactors. The inert heat carrier is used to transfer heat between the reactors. In this way, the gasification efficiency of char is improved since the tar and pyrolysis gas generated in the pyrolysis process will no longer hinder the gasification of the char. The thermodynamic equilibrium model is used to simulate the gasification process of the triple-bed circulating fluidized bed, and the sub-models are established to simulate combustion, pyrolysis, and gasification processes. The simulation results agree well with the experimental values. Besides, the model studies the effects of key parameters such as the gasification reaction temperature and the ratio of steam to C (S/C) on gasification performance. Results showed that higher gasification reaction temperature has a positive effect on gasification performance, S/C may not be too high, and excessive water vapor will directly affect the gasification reaction.
摘要:
This study investigated the passivation of multiple heavy metals (Cd, Cr, Cu, Zn, and Pb) in river sediments by composting with rice straw. The correlation between the selected environmental factors and the distribution of heavy metal fractions was assessed by multivariate analysis. The results suggested that the composting could reduce the bioavailability of Cu, Cd, Cr, and Pb. The influences of composting on the passivation of these five heavy metals are different. The greatest passivation of heavy metals is for Cd, which its bioavailability decreased 17.72%. The bioavailability of Cu, Pb, and Cr also partly decreased. However, there was almost no effect on the bioavailability of Zn. The results of multivariate analysis revealed that the predominant parameters affecting the variation of the bioavailability of different heavy metals were different in a complex composting system with multiple heavy metal pollution. While the variation partitioning analysis indicated that water-soluble carbon (WSC) was the statistically main factor affecting the bioavailability of Cd, Cr, Cu, Zn, and Pb. It provides a theoretical reference for optimizing the co-composting process in remediation of different heavy metal contamination in sediments.
摘要:
Indoor thermal environment significantly affects human health and comfort. Although several studies have investigated this aspect in numerous buildings, a gap exists with regard to the hospital waiting areas and the thermal comfort of pregnant women. Therefore, this study assessed the ambient conditions in the waiting areas of obstetric departments in four Guangzhou hospitals during summer and winter and conducted a questionnaire survey on pregnant women and males & non-pregnant female patients in the waiting areas to analyze the differences in their thermal sensation. A total of 2594 questionnaire responses were collected during the survey period, including 1477 and 1117 responses during summer and winter, respectively. Further, the predicted mean vote (PMV) model was applied to evaluate the thermal comfort and differences in the thermal sensation between pregnant women and escorts. The results indicated that the PMV model could not accurately predict the thermal comfort of the subjects and overestimated their thermal sensation. Additionally, the acceptable operative temperature ranges in the obstetric waiting areas were 23.8-28.6 degrees C and 18.5-23.5 degrees C, and acceptable standard effective temperature ranges were 21.0-27.9 degrees C and 19.0-25.8 degrees C during summer and winter, respectively. Further, the neutral temperature of pregnant women was lower than that of the males & non-pregnant female patients. The results of this study can guide the indoor thermal environments in the hospitals for the improved comfort of pregnant women, which in turn, will benefit their health as well as that of their babies.
作者机构:
[Huang, Changjun; Zhou, Qingshan; Cao, Yuanzhi] Hunan City Univ, Sch Municipal & Surveying Engn, Yiyang 413000, Peoples R China.;[Huang, Changjun] Key Lab Key Technol Digital Urban Rural Spatial P, Yiyang 413000, Peoples R China.;[Zhou, Lv] Guilin Univ Technol, Sch Geomat & Geoinformat, Guilin 541004, Peoples R China.
通讯机构:
[Changjun Huang] S;School of Municipal and Surveying Engineering, Hunan City University, Yiyang, China<&wdkj&>Key Laboratory of Key Technologies of Digital Urban-Rural Spatial Planning of Hunan Province, Yiyang, China
摘要:
Landslide is a global environmental geological hazard caused by natural or human activities. Since the impoundment of the Three Gorges reservoir area, the geological disasters such as collapses, landslides and other kinds of geological disasters increased obviously due to the periodic fluctuation of the water level in the Yangtze River. Wanzhou District is located in the center of the Three Gorges Reservoir Area, which plays an important role in the prevention and control of geological hazards in the whole Three Gorges Reservoir Area. This is because a large number of deep bedrocks old landslides are distributed among this region, such as Taibaiyan ancient landslide, Caojiezi ancient landslide, Anlesi ancient landslide, Pipaping ancient landslide, and Diaoyanping ancient landslide. In this study, Quasi-Persistent Scatterers InSAR (QPS-InSAR) time-series method is proposed to identify and monitor the ancient landslides in Wanzhou. In this method, the High-coherent test is applied to Quasi-Persistent Scatterers (PSC) selection, and PSC and persistent scatterer are combined to improve the density of measurement points in vegetation area. The QPS-InSAR method is also characterized by the appropriate combination of differential interferograms produced by a Minimum Spanning Tree and the employment of the phase triangulation algorithm to estimate the optimal phase. This technique was performed on 8 scenes of L-band ALOS PALSAR ascending data acquired during 2015-2018, then deformation rate maps and time series for ancient landslide were generated, which were applied to retrieve time series displacement for the large-scale landslide in Wanzhou District. The experiment results show that there are obvious landslide deformation patterns detected in this region with displacement velocity larger than - 21 mm/yr during the observation period. Finally, the influencing factors such as geological conditions, distribution of rainfall and reservoir water level change in the Three Gorges Reservoir area, and deformation mechanism of Wanzhou landslide are analyzed. The monitoring results will help the local government to carry out regular landslide inspection and strengthen landslide disaster early warning.