Integrate the original face image and its mirror image
for face recognition
Yong Xu a,b,Xuelong Li c,n,Jian Yang d,David Zhang e
a Bio-Computing Research Center,Shenzhen Graduate School,Harbin Institute of Technology,Shenzhen518055,Guangdong,P.R.China
b Key Laboratory of Network Oriented Intelligent Computation,Shenzhen518055,Guangdong,P.R.China
c Center for OPTical IMagery Analysis an
d Learning(OPTIMAL),Stat
e Key Laboratory o
f Transient Optics and Photonics,Xi'an Institute of Optics and Precision
Mechanics,Chinese Academy of Sciences,Xi'an710119,Shaanxi,P.R.China
d Colleg
e o
f Computer Science and Technology,Nanjin
g University of Science and Technology,Nanjing,210094,China
e Biometrics Research Center,The Hong Kong Polytechnic University,Hong Kong
a r t i c l e i n f o
Article history:
Received3June2013
Received in revised form
21September2013
Accepted22October2013
Communicated by D.Zhang
Keywords:
Pattern recognition
Biometrics
Face recognition
Mirror image
Sparse representation
a b s t r a c t
The face almost always has an axis-symmetrical structure.However,as the face usually does not have an
absolutely frontal pose when it is imaged,the majority of face images are not symmetrical images.These
facts inspire us that the mirror image of the face image might be viewed as a representation of the face
with a possible pose opposite to that of the original face image.In this paper we propose a scheme to
produce the mirror image of the face and integrate the original face image and its mirror image for
representation-based face recognition.This scheme is simple and computationally ef?cient.Almost all
the representation-based classi?cation methods can be improved by this scheme.The underlying
rationales of the scheme are as follows:?rst,the use of the mirror image can somewhat overcome the
misalignment problem of the face image in face recognition.Second,it is able to somewhat eliminate the
side-effect of the variation of the pose and illumination of the original face image.The experiments show
that the proposed scheme can greatly improve the accuracy of the representation-based classi?cation
methods.The proposed scheme might be also helpful for improving other face recognition methods.
&2013Elsevier B.V.All rights reserved.
1.Introduction
As we know,the main challenges of face recognition are that
the face image might severely vary with the various poses,facial
expression and illumination[1–3].A face recognition method
greatly suffers from these challenges.In order to address these
challenges,people have made many efforts.For example,Jian et al.
proposed the illumination compensation method for face recogni-
tion[4].Sharma et al.proposed pose invariant virtual classi?ers
for face recognition[5].We also note that if the available training
samples of a face can suf?ciently show possible variations of the
pose,facial expression and illumination,it will be possible to
obtain a high accuracy.Unfortunately,in real-world applications
a face usually has only a very small number of training samples,
which cannot convey many variations of the face[6–10].In order
to overcome the problem that the training samples of a face do not
convey suf?cient variations of the face,previous literatures
have proposed some approaches to generate new(i.e.virtual or
synthesized)face images and to enlarge the size of the set of the
training samples.
It is known that both the facial structure and the facial expression
are symmetrical[11].Previous literatures have successfully exploited
the symmetrical structure of the face for face detection[11–14].
However,it should be pointed out that in real-world face recognition
applications,a large number of face images are not symmetrical
images due to non-frontal and non-neutral pose[15].Xu et al.
proposed an approach to generate“symmetrical”face images and
exploited both the original and“symmetrical”face images to
recognize the subject[15].As the“symmetrical”face image is
generated with the assumption that the facial structure is symme-
trical,it is an axis-symmetrical image.However,as shown later,the
“symmetrical”face images obtained in[15]is not a natural face
image and even appear to be strange.
A real-world face recognition application also often suffers
from the misalignment problem of the face image.This problem
certainly makes the face image not symmetrical and is advanta-
geous for correctly recognizing the face.Fig.3presented later
shows an example of misalignment of the face image.
It should be pointed out that though previous literatures have
made many efforts in making virtual or synthesized face images
which re?ect the variation of the face as much as possible,almost
Contents lists available at ScienceDirect
journal homepage:https://www.doczj.com/doc/4f5619375.html,/locate/neucom
Neurocomputing
0925-2312/$-see front matter&2013Elsevier B.V.All rights reserved.
https://www.doczj.com/doc/4f5619375.html,/10.1016/j.neucom.2013.10.025
n Corresponding author.
E-mail addresses:yongxu@https://www.doczj.com/doc/4f5619375.html,(Y.Xu),xuelong_li@https://www.doczj.com/doc/4f5619375.html,(X.Li),
csjianyang@https://www.doczj.com/doc/4f5619375.html,(J.Yang),csdzhang@https://www.doczj.com/doc/4f5619375.html,.hk(D.Zhang).
Neurocomputing?(????)???–???
no literature generated virtual or synthesized face images by exploiting the special nature of the face i.e.the symmetry of the structure.This motivates us to exploit the symmetrical structure of the face to improve previous face recognition methods.
In this paper,we propose a novel scheme to improve the face recognition method.The proposed scheme ?rst generates the mirror image of the original face image and then applies a representation-based classi ?cation (RBC)method to both the origi-nal face image and its mirror image to perform face recognition.We also refer to the mirror image of the original face image as face mirror image.The rationales of the proposed scheme are as follows:?rst,the face mirror image re ?ects some possible change in pose and illumination of the original face image.For example,if two original face images of a same subject have a left tilt pose and right tilt pose,then the difference between them will be great in terms of the distance metric.If these two face images from the same face are used as training and test samples,then the test sample will be hard to be correctly classi ?ed.The example shown in Fig.1clearly illustrates this.As a consequence,the face recognition method using only the original face images will be very hard to obtain satisfactory accuracy.However,the difference between either of the two original face images and the mirror image of the other original face image might be very little in terms of the distance metric.As a result,the use of the face mirror image will be very useful for correctly classifying the test sample.For detailed demonstration,refer to Section 5.Second,the face mirror image is very useful to overcome the possible misalignment problem of the original face image (for detail refer to Section 5).The results of various experi-ments on face recognition also show that the proposed scheme is quite feasible and can improve the state-of-the-art representation-based classi ?cation methods.
This paper has the following main contributions.First,it proposes the scheme to integrate the original face images and their mirror images for representation-based face recognition.It also describes the rationale of the proposed scheme.Second,it shows that a number of representation-based classi ?cation methods can be improved by using the proposed scheme.
The remainder of the paper is organized as follows.Section 2presents related works.Section 3presents representation based classi ?cation methods.Section 4describes our proposed scheme.Section 5describes the rationales of the proposed scheme.Section 6shows the experimental results and Section 7offers the conclusion.
2.Related works
Because our proposed scheme is based on representation-based classi ?cation (RBC)methods and the generated virtual face images,this section mainly reviews the literatures of generating virtual or synthesized face images and those literatures on RBC
methods.A number of previous works focus on generating virtual or synthesized face images and on enlarging the size of the set of the training samples.For example,Tang et al.[16]obtained “virtual ”facial expression by exploiting the prototype faces and optic ?ow.Jung et al.[17]obtained new samples of the face by using the noise.Thian et al.[18]used simple geometric transfor-mations to make virtual samples.Ryu et al.[19]exploited the distribution of the training samples to produce virtual training samples of the face.Sharma et al.[5]extended training samples by generating multiple virtual views of a person under different poses and illumination from a single face image.Beymer et al.[6]and Vetter et al.[7]also addressed this issue by generating new samples with virtual views.
Among a variety of face recognition methods,the representation-based classi ?cation (RBC)method can achieve a very high accuracy and has received much attention [20–23].The conventional RBC method is also referred to as sparse representation classi ?cation (SRC)method [20–22].RBC including SRC assumes that the test sample can be well represented by a linear combination of all the training samples.SRC obtains its solution using the constraint of the ?1norm minimization.In other words,SRC achieves its solution with the sparsity constraint which assumes that a number of the coef ?cients of the linear combination are equal or close to zeroes.We also say that SRC uses a sparse linear combination of all the training samples to represent the test sample.
Besides SRC,RBC can be also implemented by using the constraint of the ?2norm minimization [24–26].The corresponding method can be referred to as RBC with the ?2norm minimization constraint.It seems that the algorithm of RBC with the ?2norm minimization constraint is simpler and easier to implement than that of SRC.There are two kinds of RBC with the ?2norm minimization constraint.The ?rst kind exploits the training samples from all the classes to represent the test sample and uses the representation result to perform classi ?cation [24–26].The sparsity constraint is imposed on only a few methods of this kind.For example,the methods proposed by Shi et al.[23]and by Zhang et al.[25]have no sparsity constraint.In other words,these two methods do not require that the coef ?cients of the linear combination to represent the test sample are equal or close to zeroes.On the other hand,the sparsity constraint is imposed on the method proposed in [26]in a special way.The second kind exploits the training samples from each class to represent the test sample and the classi ?cation is also performed in terms of the representation result [27,28].The sparsity constraint is also not imposed on this kind of method.Specially,this kind of method usually assumes that the test sample can be respectively approximately represented by a linear combination of the training samples of every class and no any constraint sparsity is imposed on the coef ?cients of the linear combination.A typical example of this kind of method is linear regression classi ?cation (LRC)[27].The sparse representation has also been used to other problems such as sparse graph based image annotation [29],super-resolution reconstruction [30],image alignment [31]and image de-noising [32].The sparsity has also been used to modify discriminant analysis [33,34]and locality preserving projection [35,36].The sparse graph is also used in video semantic annotation [36]and representative image selection [37,38].For recent advances and more applications of RBC,refer to literatures [39–41].
3.Representation based classi ?cation (RBC)
In this section,we introduce representation based classi ?cation (RBC)brie ?y.Because LRC has a distinct characteristic,we describe it in Section 3.1and present other RBCs in Section 3.2.We assume that there are c classes and each class has n training samples in the form of column vectors.Let x 1;…;x N be all the N training samples in the form of column vectors (N ?cn ).Column vector x ei à1Tn t
k
Fig.1.An original face image with a left tilt pose (a)and another original face image right tilt pose (b)of a same subject.It is clear that these two face images have great difference in terms of the distance metric.
Y.Xu et al./Neurocomputing ?(????)???–???
2
(k ?1;…;n )stands for the k -th training sample of the i -th subject,i ?1,2,…,c .Let column vector z stand for the test sample.3.1.LRC [27]
In this subsection we present the algorithm of LRC as follows.LRC establishes an equation for each class.The equation of the i -th class is z ?X i A i
e1T
where A i ??a i 1…a i
n ,X i ??x ei à1Tn n t1…x i n n .The solution of Eq.(1)is obtained using
^A i ?eX T i X i Tà1X T i
z e2T
The deviation between the i -th class and the test sample is de ?ned
as d i ?J z àX i ^A i J .If k ?arg min i d i
,then the test sample is assigned to the k -th class.
3.2.Brief introduction to other RBCs
For the representation based classi ?cation methods,all the methods except for LRC ?rst exploit a linear combination of all the training samples to represent the test sample.
RBC with the ?2norm minimization constraint can be brie ?y described as follows.We ?rst take collaborative representation classi-?cation (CRC)as an example to show the basic characteristics of this kind of method.CRC assumes that Eq.(3)is approximately satis ?ed z ?XB
e3T
where B ??b 1…b N T ,X ??x 1…x N .The solution of Eq.(3)is usually obtained using ^B
?eX T X tμI Tà1X T z e4T
μis a small positive constant and I is the identity matrix.Let
^B ??^b 1…^b N
T .Of course,if X T X is not singular,the solution of Eq.(3)can be also obtained using
^B
?eX T X Tà1X T z e5T
CRC calculates the residual of the test sample with respect to the i -th
class using r i ?J z àX i ^B i J where X i ??x ei à1Tn n t1…x i n n and ^B i ??^b ei à1Tn n t1…^b i n n T .If k ?arg min i r i ,then CRC assigns the test
sample to the k -th class.
The main difference between CRC and the other RBCs with the ?2norm minimization constraint is that other RBCs might have extra constraints or steps.For example,the improvement to the nearest neighbor classi ?er (INNC)[42]has the same equation and solution scheme as CRC but uses a simpler classi ?er.The two-phase sparse representation (TPSR)method [24]has the same ?rst step as CRC but exploits an extra step to obtain a sparse linear combination of all the training samples to represent the test sample.
SRC,i.e.RBC with the ?1norm minimization constraint can be brie ?y described as follows.SRC attempts to solve the following problem:min B
J B J 1;
s :t :J z àXB J 2r ε
e6T
where ε40is a constant.SRC has no closed solution and should be iteratively solved.The original SRC algorithm is very computa-tionally inef ?cient and recently some ef ?cient algorithms for SRC have been proposed [43,44].
4.The proposed scheme
The main motivation of our proposed scheme is to use a simple way to obtain more training samples and to improve the face recognition accuracy.Though the used mirror image is simply
generated from the original face image,it also appears to be a
natural image and properly re ?ects possible variation of the original face image in pose and illumination.Moreover,the mirror image is also suf ?ciently different from the original face image in terms of the distance metric,so the use of the mirror image does enable the face recognition method to exploit more available information of the face.In this section we describe the proposed scheme in detail.The proposed scheme works as follows.It ?rst generates the mirror image of each original face image for training.It then exploits both the original face image of a face for training and the corresponding mirror image as training samples of this face.As a result,a face has 2n training samples in total.For the i -th face,we use y 0ei à1Tn tk ek ?1;…;n Tto denote the mirror image of original face image
x 0ei à1Tn tk ,respectively.x 0
ei à1Tn tk represents the original face image matrix corresponding to column vector x ei à1Tn tk .The algorithm of the proposed scheme can be presented as follows:
Step 1:Suppose that the original face image matrix has P rows
and Q columns.The mirror image of an original face image has the same size.For the i -th face,mirror image
y 0ei à1Tn tk ek ?1;…;n Tis generated using y 0
ei à1Tn tk ep ;q T?
x 0ei à1Tn tk ep ;Q àq t1T,p ?1,…,P ,q ?1,…,Q .x 0
ei à1Tn tk ep ;q Tand y 0ei à1Tn tk ep ;q Tdenote the pixels located in the p -th row and q -th column of x 0ei à1Tn tk and y 0ei à1Tn tk ,respec-tively.y 0ei à1Tn tk is then converted into a column vector and is denoted by y ei à1Tn tk .As all x ei à1Tn tk and y ei à1Tn tk act as training samples,we say that there are 2cn training samples in total.
Step 2:For the i -th face ei ?1;…;c T,let X i ??x ei à1Tn n t1…x i n n
y ei à1Tn t1…y i n n .Actually,X 1;…;X c are respectively the matrices consisting of all training samples including the original face images and mirror images of the ?rst to the i -th faces.x ei à1Tn n t1…x i n n and y ei à1Tn n t1…y i n n are all column vectors.c stands for the number of the faces i.e.number of the subjects.Then X is de ?ned as X ??X 1…X c .A RBC method (LRC or an ordinary RBC method)is applied to each test sample and corresponding training samples.For an ordinary RBC method,the equation on test sample
z is expressed as z ?XB and the solution is denoted by ^B
.If the applied method is LRC,then test sample z has c equations and the i -th equation is expressed as z ?X i A i .
Let ^A i denote the solution of z ?X i A i
.Step 3:For LRC,after solutions ^A 1;…;^A c are obtained,the devia-tion between the test sample and the i -th subject is
calculated using d i ?J z àX i ^A i J .If k ?arg min i d i
,then LRC assigns the test sample to the k -th class.For other
RBC methods,solution ^B
is a vector and has 2cn entries.It is clear that each entry of ^B
is associated with one column of X (i.e.one training sample).We refer to an entry of ^B as coef ?cient of the corresponding training sample.Let ^B
i be the vector consisting of 2n entries of ^B
i.e.the coef ?cients of x ei à1Tn n t1;…;x i n n ,y ei à1Tn n t1;…;y i n n .In other words,^B
i is associated with the i -th subject.The residual of the test sample with respect to the i -th class is calculated using
r i ?J z àX i ^B
i J .If j ?arg min i r i ,then test sample z is assigned to the j -th subject.
5.Rationales of the proposed scheme
5.1.Intuitive rationales of the proposed scheme
The proposed scheme has the following rationales:?rst,it is able to reduce the side-effect of the variation of the pose and
Y.Xu et al./Neurocomputing ?(????)???–???3
illumination of the original face image.Second,it can somewhat overcome the misalignment problem of the face image in face recognition.These rationales will be demonstrated in detail below.
We ?rst show an example in which the proposed scheme can somewhat overcome the misalignment problem of the face image.Fig.2shows photos of three faces with a homogeneous background from the SCface database [45,46].Fig.3shows test samples and training samples cropped from the photos shown in Fig.2as well as the mirror images of the training samples.They have the same size and are cropped from the same original photos.For the test and training samples in the same column,the face is the same but the test sample is a shift of the training sample.This means that there exists the misalignment problem of the face.We see from Table 1that the Euclidean distance between the original test and training samples of the same subject,referred to as original distance of the same subject,is relatively larger.However,as also shown in Table 1,the Euclidean distance between the test sample and the mirror image of the training sample of the same subject,referred to as mirror distance of the same subject,is smaller.For two face images x 0and y 0,we ?rst convert them into column vectors x and y and
calculate their Euclidean distance using
????????????????????????????
ex ày TT ex ày Tq .We also see
from Fig.3that for the same face the mirror image looks more similar with the test sample and will have a smaller distance from the test sample in comparison with the original training sample.This clearly illustrates that the use of the mirror image of the original face image is helpful for a “distance ”based classi ?er to correctly classify the test sample and is able to reduce the side-effect of the mis-alignment problem of the original face image.The following example shows that the proposed scheme can somewhat overcome the side-effect of the variation of the pose
of
Fig.2.Three photos with a homogeneous
background.
Fig.3.Test samples and training samples cropped from the photos shown in Fig.2as well as the mirror images of the training samples.The ?rst and second rows show the test samples and training samples,respectively.The third row shows the mirror images of the training samples.The images in the same column are generated from a same face.
Table 1
The original and mirror distances of the samples shown in Fig.3.Each sample has the size of 1400?1200.No.of the subject 123Original distance (?105) 1.97 2.01 2.23Mirror distance (?105)
1.91
1.95
1.95
Y.Xu et al./Neurocomputing ?(????)???–???
4
the original face image.Fig.4shows six test samples and six training samples from the SCface database [45,46]as well as the mirror images of these training samples.They have the same size but have different poses.Table 2shows the original and mirror distances of the same subject.We see that for the same face the mirror distance is greatly smaller than the original distance.This implies that the simultaneous use of both the original face image and the mirror image will enable us to more accurately classify the face image.
The following example shows that the proposed scheme can somewhat overcome the variation of the illumination of the original face image.Fig.5shows several test samples and training samples of the same face from the Yale B database shown in Section 6.3.They have the same size but different illuminations.For the test sample,the right face has stronger illumination than the left face.For the training sample,the right face has weaker illumination than the left face.Table 3shows the original and mirror distances on the samples shown in Fig.5.The mirror distance is also much smaller than the original distance.This again implies that the use of the mirror image is bene ?cial for correctly recognizing the face.
5.2.More analyses of the proposed scheme
In this section,we will give more interpretation of the pro-posed scheme.As we know,RBC exploits the deviation or residual of each class to classify the test sample.RBC assigns the test sample to the class with the minimum deviation or residual.The deviation and residual indeed somewhat re ?ect the ability,to well represent the test sample,of the training samples of a class.The smaller the deviation or residual is,the greater the ability to well represent the test sample the corresponding class has.
Fig.6shows the deviations between a test sample and all the classes of the ORL face database.The deviations obtained using both LRC and the improvement to LRC (i.e.the integration of our proposed scheme and LRC)are shown in Fig.6.This test sample is from the ?fth class.From Fig.6,we see that LRC will lead to erroneous classi ?cation of the test sample,since the correspond-ing deviation between the test sample and the ?fth class is not the smallest.However,the improvement to LRC will obtain the correct classi ?cation result,because the corresponding deviation between the test sample and the ?fth class is the smallest.Fig.6also implies that the improvement to LRC has stronger ability to represent the test sample than LRC.
Fig.7shows the deviations between a test sample and all the classes of the FERET face database.This test sample is from the nineteenth class.From Fig.7,we see that LRC will lead to erroneous classi ?cation for this test sample,but the improvement to LRC will obtain the correct classi ?cation result.Shan et al.[8]also exploited the mirror image of the face image in recognizing the face,but the performance of his method is associated with the size of the blocks generated from the original face image.
The rationale of the proposed scheme can also be presented from the viewpoint of numerical analysis.For simplicity of presentation,we just take LRC and the improvement to LRC as an example.Suppose that test sample is from the i -th class.As the improvement to LRC has more available training samples than LRC,it is easy to know that the deviation between the i -th class and the test sample obtained using the improvement to LRC is usually smaller than that obtained using LRC.
https://www.doczj.com/doc/4f5619375.html,parison of our scheme with the one proposed in [15]In this subsection,we show the difference between our scheme and the scheme proposed in [15].The scheme proposed in [15]is able to obtain a virtual axis-symmetrical face image.As shown in [15],these virtual face images can somewhat re ?ect possible variation in pose and scale of the face.However,the “symmetrical ”face images obtained in [15]have the following shortcoming:they are not natural face images and even appear to be strange.Figs.8and 9show some reasonable and unreasonable “symmetrical ”face images obtained in [15].From Fig.9,we see that unreasonable “symmetrical ”face images do appear not to be natural face
images.
Fig.4.Six test samples and six training samples from the SCface database as well as the mirror images of these training samples.The ?rst and second rows show six test samples and six training samples,respectively.The third row shows the mirror images of the six training samples shown in the second row.The images in the same column are generated from a same face.
Table 2
The original and mirror distances of the samples shown in Fig.4.Each sample has the size of 75?75.No.of the subject 123456Original distance (?103)9.948.338.728.619.908.82Mirror distance (?103)
4.05
6.62
3.57
4.08
4.47
4.01
Y.Xu et al./Neurocomputing ?(????)???–???5
For example,some unreasonable “symmetrical ”face images have a very big or small nose,which not only makes the “symmetrical ”face image far from the original true face image but also looks ugly.
Differing from the virtual face image obtained using the scheme proposed in [15],the virtual face image i.e.mirror image of the original face image generated from our scheme always looks like a true face image.Another difference between the scheme proposed in [15]and our scheme is as follows.The “symmetrical ”face images obtained in [15]indeed contains much redundant information,because it is a strictly axis-symmetrical face image,the left face of which is the mirror image of the right face.In other words,one half of the pixels in the “symmetrical ”face image seem to be redundant.However,the virtual face image generated from our scheme is not an axis-symmetrical face image and does not have the same kind of redundant information as the “symmetrical ”face images obtained in [15].In addition,because our scheme generates only one virtual face image for an original face image and the scheme proposed in [15]obtains two virtual face images for an original face image,our scheme is simpler and easier to implement.
6.Experimental results
6.1.Experiments on the FERET face database
We ?rst used a subset of the FERET face database to test our method.This subset consists of 1400images from 200subjects each providing seven images [47].This subset was composed of images whose names are marked with two-character strings:‘ba ’,‘bj ’,‘bk ’,‘be ’,‘bf ’,‘bd ’,and ‘bg ’.We resized each image to a 40?40image using the down-sampling algorithm.We took the ?rst 1,2,3and 4face images of each subject as original training samples and treated the remaining face images as test samples.Table 4shows the rates of classi ?cation errors of different methods.We see that our proposed scheme can improve LRC,CRC,SRC,INNC and coarse to ?ne k nearest neighbor classi ?er (CFKNNC)[48].Hereafter,the integrations of our proposed scheme and LRC,CRC,SRC,INNC as well as KNNC are referred to as the improvements
to
Fig.5.Six test samples and six training samples from the Yale B database as well as the mirror images of these training samples.The ?rst and second rows show six test samples and six training samples,respectively.The third row shows the mirror images of the six training samples shown in the second row.The images in the same column are generated from a same face.
Table 3
The original and mirror distances on the samples shown in Fig.5.No.of the column 123456Original distance (?104) 2.47 2.16 3.23 2.30 2.07 1.36Mirror distance (?104)
2.34
2.01
3.10
2.15
1.95
1.21
Fig.6.The deviations between a test sample and all the classes of the ORL face database.The deviations are obtained by using LRC and the improvement to LRC.The test sample is from the ?fth
class.
Fig.7.The deviations between a test sample and all the classes of the FERET face database.The deviations are obtained by using LRC and the improvement to LRC.The test sample is from the nineteenth class.
Y.Xu et al./Neurocomputing ?(????)???–???
6
Fig.8.Some original face images from the FERET face database and the corresponding “symmetrical ”face images.The ?rst row shows the original face images.The second and third rows respectively show the ?rst and second symmetrical face images,of the original face images,obtained using the scheme proposed in [15].Because these “symmetrical ”face images appear to be natural face images,we refer to them as reasonable “symmetrical ”face
images.
Fig.9.Some original training samples from the AR face database and the corresponding symmetrical face images.The ?rst row shows the original face images.The second Y.Xu et al./Neurocomputing ?(????)???–???7
LRC,CRC,SRC,INNC as well as KNNC,respectively.When each subject provided two training samples,the rates of classi?cation errors of LRC,CRC,INNC and CFKNNC are35.9%,41.6%,41.7%and 36.70%,respectively.However,the rates of classi?cation errors of the improvements to LRC,CRC,INNC and CFKNNC are22.4%,34.5%, 33.5%,and27.70%,respectively.When CFKNNC and improvement to CFKNNC were implemented,we set parameter n and K to n?N/4and K?1.N stands for the number of all the original training samples.K?1means that the nearest neighbor classi?cation is indeed performed.
6.2.Experiments on the ORL face database
The ORL database[49]includes400face images taken from40 subjects each providing10face images.For some subjects,the images were taken at different times,with varying lighting,facial expressions(open/closed eyes,smiling/not smiling),and facial details(glasses/no glasses).Each image was also resized to a56 by46image matrix by using the down-sampling algorithm.We took the?rst1,2,3and4face images of each subject as original training samples and treated the remaining face images as test samples.The experimental results were shown in Table5.We see again that proposed scheme can improve all the methods.
6.3.Experiments on the Yale B face database
In this subsection we use the Yale B[50]and the Extended Yale B[51]face databases to conduct experiments.There are10 subjects in the Yale B database,and28subjects in the extended Yale B database.Each subject has64images under different illumination conditions.The facial portion of each original image was cropped to a192?168image.We resized the cropped image to a96by84matrix.In order to computationally ef?ciently perform improvement to CFKNNC and CFKNNC,we did as follows. Before improvement to CFKNNC and CFKNNC were implemented, the face image was further resized to48by42.We took the?rst 16,20and24face images of each subject as original training samples and treated the remaining face images as test samples. Table6shows the rates of classi?cation errors(%)of different methods on the Yale B database.The experimental results show again that our proposed scheme can improve all the methods. 7.Conclusions
Our this paper clearly demonstrates that the mirror image of the face image can be exploited to simulate possible variation of the face image and is able to reduce the side-effect of the pose and illumination difference between the training and test samples of the same face.The proposed scheme can also overcome the misalignment problem of the face image,which usually reduces the accuracy of face recognition.The proposed scheme is very simple and computationally very ef?cient and improves RBCs at a low computational cost.The analyses and experimental results suf?ciently show the rationales of the proposed scheme.As the proposed scheme provides a good way to exploit the special nature of the face and is helpful for achieving better recognition results,it can be also applied to other face recognition methods. Acknowledgments
This paper is partially supported by the National Basic Research Program of China(973Program)(Grant no.2012CB316400),the National Natural Science Foundation of China(Grant nos.61370163, 61233011,61203376,61332011,and61125106),Shenzhen Municipal Science and Technology Innovation Council(Nos.JC201005260122A, JCYJ20120613153352732and CXZZ20120613141657279),and the Shaanxi Key Innovation Team of Science and Technology(Grant no.: 2012KCT-04).
References
[1]L.Pishchulin,T.Gass,P.Dreuw,H.Ney,Image warping for face recognition:
from local optimality towards global optimization,Pattern Recognit.45(9) (2012)3131–3140.
[2]S.N.Kautkar,G.A.Atkinson,M.L.Smith,Face recognition in2D and2.5D using
ridgelets and photometric stereo,Pattern Recognit.45(9)(2012)3317–3327.
[3]J.Wang,J.You,Q.Li,Y.Xu,Orthogonal discriminant vector for face recognition
across pose,Pattern Recognit.45(12)(2012)4069–4079.
[4]M.Jian,https://www.doczj.com/doc/4f5619375.html,m,J.Dong,Illumination compensation and enhancement for
face recognition,in:Proceedings of Asia–Paci?c Signal and Information Processing Association Annual Summit Conference(APSIPA ASC'2011),paper Wed-AM.RS6,2011.
[5]A.Sharma,A.Dubey,P.Tripathi,V.Kumar,Pose invariant virtual classi?ers
from single training image using novel hybrid-eigenfaces,Neurocomputing73 (10–12)(2010)1868–1880.
[6]D.Beymer,T.Poggio,Face recognition from one example view,in:Proceedings
of the Fifth International Conference on Computer Vision,1995,pp.500–507.
[7]T.Vetter,Synthesis of novel views from a single face image,https://www.doczj.com/doc/4f5619375.html,put.Vis.
28(2)(1998)102–116.
Table4
The rates of classi?cation errors(%)of different methods on the FERET database.
Number of the training samples of each subject4321
LRC21.540.1335.9055.08 Improvement to LRC14.1722.1322.4050.83 INNC42.6749.5041.7056.50 Improvement to INNC35.5035.7533.5054.08 CRC44.6755.6341.6055.67 Improvement to CRC36.5036.8834.5053.25 SRC23.3340.0035.2049.75 Improvement to SRC21.0039.7534.7049.00 CFKNNC38.5045.1236.7052.17 Improvement to CFKNNC25.5029.0027.7051.17
Table5
The rates of classi?cation errors(%)of different methods on the ORL database. Number of the training samples of each subject4321
LRC13.7518.5720.6232.50 Improvement to LRC12.0812.1417.8128.89 INNC12.5017.8618.4428.33 Improvement to INNC14.1714.6416.5628.06 CRC11.5013.9316.5631.94 Improvement to CRC13.3312.1414.6928.89 SRC10.0014.2915.0027.50 Improvement to SRC8.3312.1414.0626.67 CFKNNC14.5819.2917.8126.39 Improvement to CFKNNC12.5014.2915.0026.67Table6
The rates of classi?cation errors(%)of different methods on the Yale B database.
Number of the training samples of each subject242016
LRC16.5123.3324.73 Improvement to LRC12.7620.9324.07 INNC33.2931.0431.96 Improvement to INNC23.9527.1531.03 CRC22.5724.7026.86 Improvement to CRC19.5421.7724.01 SRC40.8643.1846.27 Improvement to SRC35.2037.2037.06 CFKNNC29.2829.6131.03 Improvement to CFKNNC28.6827.5729.77
Y.Xu et al./Neurocomputing?(????)???–???8
[8]G.J.Shan,Virtual sample generating for face recognition from a single training
sample per person,Sci.Technol.Eng.13(14)(2013)3908–3911.(in Chinese).
[9]X.Tan,S.Chen,Z.-H.Zhou,F.Zhang,Face recognition from a single image per
person:a survey,Pattern Recognit.39(9)(2006)1725–1745.
[10]L.Qiao,S.Chen,X.Tan,Sparsity preserving discriminant analysis for single
training image face recognition,Pattern Recognit.Lett.31(5)(2010)422–429.
[11]P.Ekman,J.C.Hager,W.V.Friesen,The symmetry of emotional and deliberate
facial actions,Psychophysiology18(2)(1981)101–106.
[12]E.Saber, A.Murat Tekalp,Frontal-view face detection and facial feature
extraction using color,shape and symmetry based cost functions,Pattern Recognit.Lett.19(8)(1998)669–680.
[13]M.-C.Su, C.-H.Chou,Application of associative memory in human face
detection,in:International Joint Conference on Neural Networks,vol.5, 1999,pp.3194–3197.
[14]S.Saha,S.Bandyopadhyay,A symmetry based face detection technique,in:
Proceedings of the IEEE WIE National Symposium on Emerging Technologies, 2007,pp.1–4.
[15]Y.Xu,X.Zhu,Z.Li,G.Liu,Y.Lu,H.Liu,Using the original and‘symmetrical face’
training samples to perform representation based two-step face recognition, Pattern Recognit.46(4)(2013)1151–1158.
[16]B.Tang,S.Luo,H.Huang,High performance face recognition system by
creating virtual sample,in:Proceedings of International Conference on Neural Networks and Signal Processing,2003,pp.972–975.
[17]H.-C.Jung,Authenticating corrupted face image based on noise model,in:
Proceedings of the Sixth IEEE International Conference on Automatic Face and Gesture Recognition,2004,pp.272–277.
[18]N.P.H.Thian,S.Marcel,S.Bengio,Improving face authentication using virtual
samples,in:Proceedings of the IEEE International Conference on Acoustics, Speech,and Signal Processing,vol.3,2003,pp.III-233–236.
[19]Y.-S.Ryu,S.-Y.Oh,Simple hybrid classi?er for face recognition with adaptively
generated virtual data,Pattern Recognit.Lett.23(7)(2002)833–841.
[20]J.Wright,Y.Ma,J.Mairal,et al.,Sparse representation for computer vision and
pattern recognition,Proc.IEEE98(6)(2010)1031–1044.
[21]J.Wright, A.Y.Yang, A.Ganesh,et al.,Robust face recognition via sparse
representation,IEEE Trans.Pattern Anal.Mach.Intell.31(2)(2009)210–227.
[22]J.Yang,L.Zhang,Y.Xu,J.-Y.Yang,Beyond sparsity:the role of L1-optimizer in
pattern classi?cation,Pattern Recognit.45(3)(2012)1104–1118.
[23]Y.Shi,D.Q.Dai,C.C.Liu,H.Yan,Sparse discriminant analysis for breast cancer
biomarker identi?cation and classi?cation,Prog.Natl.Sci.19(11)(2009) 1635–1641.
[24]Y.Xu,D.Zhang,J.Yang,J.-Y.Yang,A two-phase test sample sparse representa-
tion method for use with face recognition,IEEE Trans.Circuits Syst.Video Technol.21(9)(2011)1255–1262.
[25]L.Zhang,et al.,Sparse representation or collaborative representation:which
helps face recognition?in:Proceedings of the ICCV,2011.
[26]Y.Xu,Q.Zhu,Z.Fan,D.Zhang,J.Mi,https://www.doczj.com/doc/4f5619375.html,i,Using the idea of the sparse
representation to perform coarse to?ne face recognition,Inf.Sci.238(2013) 138–148.
[27]I.Naseem,R.Togneri,M.Bennamoun,Linear regression for face recognition,
IEEE Trans.Pattern Anal.Mach.Intell.32(11)(2010)2106–2112.
[28]I.Naseem,R.Togneri,M.Bennamoun,Robust regression for face recognition,
Pattern Recognit.45(1)(2012)104–118.
[29]J.Tang,R.Hong,S.Yan,T.-S.Chua,et al.,Image annotation by kNN-sparse
graph-based label propagation over noisily tagged web images,ACM Trans.
Intell.Syst.Technol.2(2)(2011)14.
[30]J.Yang,J.Wright,T.S.Huang,Y.Ma,Image super-resolution as sparse
representation of raw image patches,in:CVPR2008,2008,pp.1–8.
[31]A.Wagner,J.Wright, A.Ganesh,Z.Zhou,et al.,Toward a practical face
recognition system:robust alignment and illumination by sparse representa-tion,IEEE Trans.Pattern Anal.Mach.Intell.34(2)(2012)372–386.
[32]W.Dong,X.Li,L.Zhang,G.Shi,Sparsity-based image denoising via dictionary
learning and structural clustering,in:CVPR2011,2011,pp.457–464.
[33]https://www.doczj.com/doc/4f5619375.html,i,W.K.Wong,Z.Jin,J.Yang,Y.Xu,Sparse approximation to the
eigensubspace for discrimination,IEEE Trans.Neural Netw.Learn.Syst.23
(12)(2012)1948–1960.
[34]https://www.doczj.com/doc/4f5619375.html,i,M.Wan,Z.Jin,J.Yang,Sparse two-dimensional local discriminant
projections for feature extraction,Neurocomputing74(4)(2011)629–637. [35]L.Qiao,S.Chen,X.Tan,Sparsity preserving projections with applications to
face recognition,Pattern Recognit.43(1)(2010)331–341.
[36]Z.-J.Zha,M.Wang,Y.-T.Zheng,et al.,Interactive video indexing with statistical
active learning,IEEE Trans.Multimed.14(1)(2012)17–27.
[37]Z.-J.Zha,L.Yang,T.Mei,Visual query suggestion,in:Proceedings of ACM
International Conference on Multimedia,2009,pp.15–24.
[38]Z.-J.Zha,L.Yang,T.Mei,Visual query suggestion:towards capturing user
intent in internet image search,ACM https://www.doczj.com/doc/4f5619375.html,mun.
Appl.6(3)(2010),DOI:https://www.doczj.com/doc/4f5619375.html,/10.1145/1823746.1823747.
[39]M.Yang,L.Zhang,J.Yang, D.Zhang,Regularized robust coding for face
recognition,IEEE Trans.Image Process.(2013),arXiv:1202.4207.
[40]X.Gao,N.Wang,D.Tao,X.Li,Face sketch-photo synthesis and retrieval using
sparse representation,IEEE Trans.Circuits Syst.Video Technol.22(8)(2012) 1213–1226.
[41]X.Gao,Kaibing Zhang,Dacheng Tao,Xuelong Li,Image super-resolution with
sparse neighbor embedding,IEEE Trans.Image Process.21(7)(2012) 3194–3205.[42]Y.Xu,Q.Zhu,Y.Chen,J.-S.Pan,An improvement to the nearest neighbor
classi?er and face recognition experiments,https://www.doczj.com/doc/4f5619375.html,put.Inf.Control9
(2)(2013)543–554.
[43]C.Y.Peng,J.W.Li,Fast sparse representation model for l1-norm minimisation
problems,Electron.Lett.48(3)(2012)162–164.
[44]J.-B.Huang,M.-H.Yang,Fast sparse representation with prototypes,in:CVPR
2010,2010,pp.3618–3625.
[45]Available:?https://www.doczj.com/doc/4f5619375.html,/?(Online).
[46]M.Grgic,K.Delac,S.Grgic,SCface—surveillance cameras face database,
Multimed.Tools Appl.J.51(3)(2011)863–879.
[47]Available:?https://www.doczj.com/doc/4f5619375.html,/iad/humanid/feret/feret_master.html?(Online).
[48]Y.Xu,Q.Zhu,Z.Fan,et al.,Coarse to?ne K nearest neighbor classi?er,Pattern
Recognit.Lett.34(9)(2013)980–986.
[49]Available:?https://www.doczj.com/doc/4f5619375.html,/research/dtg/attarchive/facedatabase.html?
(Online).
[50]Athinodoros S.Georghiades,Peter N.Belhumeur,David J.Kriegman,From few
to many:illumination cone models for face recognition under variable lighting and pose,IEEE Trans.Pattern Anal.Mach.Intell.23(6)(2001)643–660. [51]Available:?https://www.doczj.com/doc/4f5619375.html,/leekc/ExtYaleDatabase/ExtYaleB.html?
(Online).
Yong Xu was born in Sichuan,China,in1972.He
received his B.S.degree and M.S.degree at Air Force
Institute of Meteorology(China)in1994and1997,
respectively.He then received his Ph.D.degree in
Pattern recognition and Intelligence System at the
Nanjing University of Science and Technology(NUST)
in2005.From May2005to April2007,he worked at
Shenzhen Graduate School,Harbin Institute of Tech-
nology(HIT)as a Postdoctoral Research Fellow.Now he
is an Associate Professor at Shenzhen graduate school,
HIT.His current interests include Pattern Recognition,
Biometrics,Image Processing and Video Analysis. Xuelong Li is a Full Professor with the Center for OPTical IMagery Analysis and Learning(OPTIMAL),State Key Laboratory of Transient Optics and Photonics,Xi'an Institute of Optics and Precision Mechanics,Chinese Academy of Sciences,Xi'an 710119,Shaanxi,P.R.
China.
Jian Yang received the B.S.degree in Mathematics from
the Xuzhou Normal University in1995,the M.S.degree
in Applied Mathematics from the Changsha Railway
University in1998and the Ph.D.degree in Computer
Science from the Nanjing University of Science and
Technology(NUST)in2002.He was a Postdoctoral
Researcher at the University of Zaragozain2003.From
2004to2006,he was a Postdoctoral Fellow in Depart-
ment of Hong Kong Polytechnic University.From2006
to2007,he was a Postdoctoral Fellow in Department of
Computer Science of New Jersey Institute of Technol-
ogy.From2007,he has been a Professor in the School of
Computer Science and Technology of NUST.He is the author of more than50scienti?c papers in Pattern Recognition,Computer Vision and the related areas.His journal papers have been cited more than1100times in the ISI Web of Science,and2400times in the Web of Scholar Google.Currently,he is an Associate Editor of Pattern Recognition Letters and IEEE Transactions on Neural
Networks.
David Zhang graduated in Computer Science from
Peking University.He received his M.Sc.in Computer
Science in1982and his Ph.D.in1985from the Harbin
Institute of Technology(HIT).From1986to1988he
was a Postdoctoral Fellow at Tsinghua University and
then an Associate Professor at the Academia Sinica,
Beijing.In1994he received his second Ph.D.in Elec-
trical and Computer Engineering from the University of
Waterloo,Ontario,Canada.Currently,he is a Chair
Professor at the Hong Kong Polytechnic University
where he is the Founding Director of the Biometrics
Technology Centre(UGC/CRC)supported by the Hong
Kong SAR Government in1998.He also serves as Visiting Chair Professor in Tsinghua University,and Adjunct Professor in Shanghai Jiao Tong University,Peking University,Harbin Institute of Technology,and the University of Waterloo.He is the Founder and Editor-in-Chief,International Journal of Image and Graphics(IJIG);Book Editor,Springer International Series on Biometrics(KISB);Organizer,the?rst International Conference on Biometrics Authentication(ICBA);Associate Editor of more than ten international journals including IEEE Transactions and Pattern Recognition;Technical Committee Chair of IEEE CIS and the author of more than10books and200journal papers.Professor Zhang is a Croucher Senior Research Fellow,Distinguished Speaker of the IEEE Computer Society,and a Fellow of both IEEE and IAPR.
Y.Xu et al./Neurocomputing?(????)???–???9
五年级上册成语解释及近义词反义词和造句大全 囫囵吞枣；【解释】：囫囵：整个儿。把枣整个咽下去，不加咀嚼，不辨味道。比喻对事物不加分析考虑。【近义词】：不求甚解【反义词】融会贯穿[造句];学习不能囫囵吞枣而是要精益求精 不求甚解；bùqiúshènjiě【解释】：甚：专门，极。只求明白个大概，不求完全了解。常指学习或研究不认真、不深入【近义词】：囫囵吞枣【反义词】：精益求精 造句；1；在学习上，我们要理解透彻，不能不求甚解 2；学习科学文化知识要刻苦钻研，深入领会，不能粗枝大叶，不求甚解。 千篇一律；【解释】：一千篇文章都一个样。指文章公式化。也比喻办事按一个格式，专门机械。 【近义词】：千人一面、如出一辙【反义词】：千差万别、形形色色 造句；学生旳作文千篇一律，专门少能有篇与众不同旳，这确实是平常旳练习太少了。 倾盆大雨；qīngpéndàyǔ【解释】：雨大得象盆里旳水直往下倒。形容雨大势急。 【近义词】：大雨如柱、大雨滂沱【反义词】：细雨霏霏牛毛细雨 造句；3月旳天说变就变，瞬间下了一场倾盆大雨。今天下了一场倾盆大雨。 坚决果断；áobùyóuyù：意思；做事果断，专门快拿定了主意，一点都不迟疑，形容态度坚决 近义词；不假思索斩钉截铁反义词；犹豫不决 造句；1看到小朋友落水，司马光坚决果断地搬起石头砸缸。2我坚决果断旳承诺了她旳要求。 饥肠辘辘jīchánglùlù【近义词】：饥不择食【反义词】：丰衣足食 造句；1我放学回家已是饥肠辘辘。2那个饥肠辘辘旳小孩差不多两天没吃饭了 滚瓜烂熟gǔnguālànshóu〔shú)【解释】：象从瓜蔓上掉下来旳瓜那样熟。形容读书或背书流利纯熟。【近义词】：倒背如流【反义词】：半生半熟造句；1、这篇课文我们早已背得滚瓜烂熟了 流光溢彩【liúguāngyìcǎi】解释；光影，满溢旳色彩，形容色彩明媚 造句：国庆节，商场里装饰旳流光溢彩。 津津有味；jīnjīnyǒuwèi解释：兴趣浓厚旳模样。指吃得专门有味道或谈得专门有兴趣。 【近义词】：兴致勃勃有滋有味【反义词】：索然无味、枯燥无味 造句；1今天旳晚餐真丰富，小明吃得津津有味。 天长日久；tiānchángrìjiǔ【解释】：时刻长，生活久。【近义词】：天长地久【反义词】：稍纵即逝 造句：小缺点假如不立即改掉, 天长日久就会变成坏适应 如醉如痴rúzuìrúchī【解释】：形容神态失常，失去自制。【近义词】：如梦如醉【反义词】：恍然大悟造句；这么美妙旳音乐，我听得如醉如痴。 浮想联翩【fúxiǎngliánpiān解释】：浮想：飘浮不定旳想象；联翩：鸟飞旳模样，比喻连续不断。指许许多多旳想象不断涌现出来。【近义词】：思绪万千 造句；1他旳话让人浮想联翩。2:这幅画饱含诗情，使人浮想联翩，神游画外，得到美旳享受。 悲欢离合bēihuānlíhé解释；欢乐、离散、聚会。泛指生活中经历旳各种境遇和由此产生旳各种心情【近义词】：酸甜苦辣、喜怒哀乐【反义词】：平淡无奇 造句；1人一辈子即是悲欢离合,总要笑口常开,我们旳生活才阳光明媚. 牵肠挂肚qiānchángguàdù【解释】：牵：拉。形容十分惦念，放心不下 造句；儿行千里母担忧，母亲总是那个为你牵肠挂肚旳人 如饥似渴rújīsìkě：形容要求专门迫切，仿佛饿了急着要吃饭，渴了急着要喝水一样。 造句；我如饥似渴地一口气读完这篇文章。他对知识旳如饥似渴旳态度造就了他今天旳成功。 不言而喻bùyánéryù【解释】：喻：了解，明白。不用说话就能明白。形容道理专门明显。 【近义词】：显而易见【反义词】：扑朔迷离造句；1珍惜时刻，好好学习，那个道理是不言而喻旳 与众不同；yǔzhòngbùtóng【解释】：跟大伙不一样。 〖近义词〗别出心裁〖反义词〗平淡无奇。造句； 1从他与众不同旳解题思路中，看出他专门聪慧。2他是个与众不同旳小孩
中医药文化是中国传统文化的瑰宝，它植根于中国传统文化的土壤之中并深受其浸染。因此，在其生长与传播的过程中，必然要打上中国传统文化的烙印。随着中国与马来西亚贸易的发展，华人移居马来西亚的规模化，以及双方文化交流的频繁，有着华人文化印记的中医药文化飘洋过海，逐渐在马来西亚扎根发展。这一方面与马来西亚特殊的气候条件和丰富的中草药资源有着很大的关系，但更重要的是马来西亚社会中相对独立发展的华人文化为其提供了近似中国国内中医药发展的文化背景，推动着马来西亚中医药文化从无到有，从区域性向全国性、从基础性向理论性、从个别性向有组织性发展。１ 中医药文化在马来西亚传播的文化背景中医药文化在马来西亚的传播，与中国传统文化的特质有着很大的关系。华人向来都有一种依恋故土的情怀，这表现在其移居他国时往往仍较多地保存着祖国的文化传统。早期移居马来西亚的华人大多是农民，他们大多在国内生活窘迫，到国外时力量单薄，倍受艰辛，因此形成一个个相对独立的的小圈子。这样一来，整个华人社会就呈现出由多个相对独立的小圈子构成的特点。“他们盲目地忠于传统，固执地保持着由他们的祖先传 下来的已有几百年历史的风俗习惯。” ［１］１５ 华人对中华传统文化执着的固守，主要体现在其组建的会馆、宗祠和庙宇中。 会馆是华人在马来西亚最早建立的地域性组织，自清嘉庆六年（ １８０１）在槟榔屿成立最早的仁和公司和广东暨汀州会馆起，华人会馆即伴随着华人社会的发展而不断发展。至１９世纪末，顺应东南亚“海外华人会馆文化”的勃兴，马来西亚华人会馆也蓬勃发展，形成了福建、广东、潮州、客家 和海南５个大帮，各帮所属会馆多达百家［２］ 。这些 会馆是移居马来西亚的华人依靠强烈的地域认同感，即包括方言、风俗、习惯在内的地域文化而建立的互济互助的组织，它们发挥着“联乡邑之盛情，谋桑梓之幸福，推而进于社会之提携、国民之结合”的作用。在这些会馆内，中华传统文化的习俗，特别是方言的同一性得到了最大程度的强化，具有相对的排他性。而这也为中医药文化在特定时期特定地点的传承，提供了一定的空间。如１８７８年，居住在吉隆坡的大埔人成立了一个团体，名叫“茶阳公司”，座落在吉隆坡的洛士街附近，作为同乡联络感情和急难互助的机构，茶阳公司附设一间留医所，供患病的同乡治病和殓葬不幸去世的同乡，内有一名中医师在留医所驻诊，这就是“回春馆中医留医所”创立的开始。 宗祠或氏族宗亲组织以血缘关系为基础，它源于中国传统的宗法社会，一方面体现了华人慎终追远，“使后嗣勿忘其之所出”的独特气质；一方面也反映了“互惠互助，敦睦亲族”的文化精神。因此，宗祠在保存宗族的伦理道德、风俗习惯等华人文化方面，具有非同寻常的意义。马来西亚华人社会中较为著名的宗祠有福建人邱华东、邱心美在槟榔屿建立的龙山堂、邱公司（１８３５）；广东华侨建立的伍氏馆（１８４８）；福建人林清甲在槟榔屿建立的林开勉述堂、林公司、敦关堂；陈金钟、陈明水的 收稿日期：２００７－０１－１８ 作者简介：郑志锋（１９７５—），男，硕士，讲师，主要从事海外华人文化研究。 马来西亚华人文化与中医药文化传承 郑志锋 （福建中医学院公共管理系，福建福州３５０１０８） 摘要：考察中医药文化在马来西亚传播的文化背景，梳理马来西亚中医药文化发展的三大途径，认为马来西亚华人文化与中医药文化之间是双向的互动关系，前者为后者提供了发展的土壤，后者则进一步丰富了前者的内涵。 关键词：马来西亚；华人文化；中医药文化；传承关系中图分类号：Ｒ０９３ 文献标识码：Ａ 文章编号：１００４－５６２７（２００７）０３－００５６－０４ ?医史研究? ＪｏｕｒｎａｌｏｆＦｕｊｉａｎＣｏｌｌｅｇｅｏｆＴＣＭＪｕｎｅ２００７，１７（３） 福建中医学院学报２００７年６月第１７卷第３期 ５６
悲惨的近义词反义词和造句 导读：悲惨的近义词 悲凉(注释:悲哀凄凉：～激越的琴声。) 悲惨的反义词 幸福(注释:个人由于理想的实现或接近而引起的一种内心满足。追求幸福是人们的普遍愿望，但剥削阶级把个人幸福看得高于一切，并把个人幸福建立在被剥削阶级的痛苦之上。无产阶级则把争取广大人民的幸福和实现全人类的解放看作最大的幸福。认为幸福不仅包括物质生活，也包括精神生活;个人幸福依赖集体幸福，集体幸福高于个人幸福;幸福不仅在于享受，而主要在于劳动和创造。) 悲惨造句 1.一个人要发现卓有成效的真理，需要千百个人在失败的探索和悲惨的错误中毁掉自己的生命。 2.贝多芬的童年尽管如是悲惨，他对这个时代和消磨这时代的地方，永远保持着一种温柔而凄凉的回忆。 3.卖火柴的小女孩在大年夜里冻死了，那情景十分悲惨。 4.他相信，他们每个人背后都有一个悲惨的故事。 5.在那次悲惨的经历之后，我深信自己绝对不是那种可以离家很远的人。 6.在人生的海洋上，最痛快的事是独断独航，但最悲惨的却是回头无岸。 7.人生是艰苦的。对不甘于平庸凡俗的人那是一场无日无夜的斗
争，往往是悲惨的、没有光华的、没有幸福的，在孤独与静寂中展开的斗争。……他们只能依靠自己，可是有时连最强的人都不免于在苦难中蹉跎。罗曼·罗兰 8.伟大的心胸，应该表现出这样的气概用笑脸来迎接悲惨的厄运，用百倍的勇气来应付开始的不幸。鲁迅人在逆境里比在在顺境里更能坚强不屈。遇厄运时比交好运时容易保全身心。 9.要抓紧时间赶快生活，因为一场莫名其妙的疾病，或者一个意外的悲惨事件，都会使生命中断。奥斯特洛夫斯基。 10.在我一生中最悲惨的一个时期，我曾经有过那类的想法：去年夏天在我回到这儿附近的地方时，这想法还缠着我;可是只有她自己的亲自说明才能使我再接受这可怕的想法。 11.他们说一个悲惨的故事是悲剧，但一千个这样的故事就只是一个统计了。 12.不要向诱惑屈服，而浪费时间去阅读别人悲惨的详细新闻。 13.那起悲惨的事件深深地铭刻在我的记忆中。 14.伟大的心胸，应该用笑脸来迎接悲惨的厄运，用百倍的勇气来应付一切的不幸。 15.一个人要发现卓有成效的真理，需要千百万个人在失败的探索和悲惨的错误中毁掉自己的生命。门捷列夫 16.生活需要爱，没有爱，那些受灾的人们生活将永远悲惨;生活需要爱，爱就像调味料，使生活这道菜充满滋味;生活需要爱，爱让生活永远充满光明。
改善服务门诊一站式服务平台实施方案 文稿归稿存档编号：[KKUY-KKIO69-OTM243-OLUI129-G00I-FDQS58-
优化门诊服务实施方案一、目的： 优化服务流程，切实改善患者就医感受，缩短排队等候时间，和谐医患关系，不断提升医患满意度。 二、实施方向： 成立门诊服务中心“一站式服务平台”。 三、具体方案如下： 1．组织管理：医院一站式服务接待中心由门诊办公室统一管理，负责人员培训；护理部协助人员调配；计算机中心协助信息化维护；后勤部协助服务台建设。 2．服务项目：预检分诊、传染病筛查、预约挂号（取号）、健康咨询、投诉接待、发放检验报告、提供便民服务，接待双向转诊上转患者。3．服务时间：上午7:30~下午17:00，双休日、节假日照常服务。4．接待中心设立在门诊一楼显眼位置，服务电话xx。 四、具体工作内容： 1.工作人员由门诊部按服务项目合理安排，服务到位。对不能立即落实 的，进行登记，并与相关部门联系，及时答复。 2.门诊代理主任现场值班，审核签章工作前移。 3.预检分诊，指引患者选择对应科室就诊，避免患者挂错科室，延误患 者就诊，减少患者退改号例数。
4.提供患者复诊预约服务，并协调相关科室安排就诊时间，指导患者自 助挂号。 5.为患者提供健教处方、专家简介、特色科室简介等资料；提供健康咨 询。 6.设意见箱，发放门诊满意度调查问卷，指导患者就诊结束后将问卷投 入意见箱。 7.接待门诊投诉，当日解决患者投诉问题。职能范围内不能解决的问题 转交投诉办公室。 8.指导患者使用检验报告自助打印机，为无自助能力的患者打印检验报 告。 9.拓展便民服务，为85岁以上患者提供轮椅，全程陪诊。 10.设立巡导志愿者岗位，安排导医、志愿者协助年老、行动不便患者办 理就诊手续。 11.接待新农合、社区转诊患者，安排就诊。 五、服务要求与管理 1.遵循“以病人为中心”的服务宗旨，落实服务接待中心的各项制度。 2.坚持规范服务，统一着装，挂牌上岗。 3.耐心、细致地回答病人的各种询问。 4.按要求认真做好工作记录。 5.服从工作安排，坚守岗位，不迟到早退，严禁脱岗。 6.巡导志愿者每两小时轮换一次。 八、设备与环境配备
西亚概况 中国网| 时间：2005-05-16 | 文章来源：外交部网站 （最近更新时间：2004年10月18日） 【国名】马来西亚（Malaysia）。 【面积】330257平方公里（马来西亚财政部2002/2003年度经济报告）。 【人口】2452.7万（2002年）。其中马来人及其他原住民占66.1%，华人占25.3%，印度人占7.4%。沙捞越州原住居民中以伊班族为主，沙巴州以卡达山族为主。马来语为国语，通用英语，华语使用也较广泛。伊斯兰教为国教，其他宗教有佛教、印度教和基督教等。 【首都】吉隆坡（Kuala Lumpur），人口约129.8万。 最高元首端古·赛义德·西拉杰丁(TUANKU SYED SIRAJUDDIN Ibni Almarhum Tuanku Syed Putra Jamalullail)，2001年12月13日就任第12任最高元首，2002年4月25日举行登基仪式。 【重要节日】马来西亚节日很多，全国大大小小的节日约有上百个。但政府规定的全国性节日只有10个，其中除少数有固定日期外，其余的具体日期由政府在前一年统一公布。主要节日有：元旦、开斋节（穆斯林）、春节（华人）、哈芝节（穆斯林）、屠妖节（印度人）、“五·一”节、圣诞节、卫塞节、现任最高元首诞辰。8月31日为马来西亚国庆（又称独立日）。 【简况】位于东南亚，国土被南中国海分隔成东、西两部分。西马位于马来半岛南部，北与泰国接壤，南与新加坡隔柔佛海峡相望，东临南中国海，西濒马六甲海峡。东马位于加里曼丹岛北部，与印尼、菲律宾、文莱相邻。全国海岸线总长4192公里。属热带雨林气候。内地山区年均气温22℃-28℃，沿海平原为25℃-30℃。 公元初马来半岛有羯荼、狼牙修等古国。15世纪初以马六甲为中心的满剌加王国统一了马来半岛的大部分。16世纪开始先后被葡萄牙、荷兰、英国占领。20世纪初完全沦为英国殖民地。沙捞越、沙巴历史上属文莱，1888年两地沦为英国保护地。二次大战中，马来亚、沙捞越、沙巴被日本占领。战后英国恢复殖民统治。1957年8月31日马来亚联合邦宣布独立。1963年9月16日，马来亚联合邦同新加坡、沙捞越、沙巴合并组成马来西亚（1965年8月9日新加坡退出）。 【政治】1999年11月，马来西亚举行第十届全国大选。执政党联盟国民阵线获超过三分之二多数席位，继续执政，马哈蒂尔第五次蝉联总理。反对党伊斯兰教党除原有的吉兰丹州外，还夺得了丁加奴州政权。2000年，马政局总体保持稳定，主要执政党巫统顺利举行了换届选举。2001年，受“9·11事件”冲击，反对党伊斯兰教党因被指责与宗教极端势力有牵连而陷于被动。马哈蒂尔总理领导的国民阵线在沙捞越州选举和玻璃市等州的补选中大获全胜，执政地位进一步巩固。2002年6月，马哈蒂尔总理宣布将于2003年10月辞去党政一切职务，并明确表示由巴达维副总理接班。马来西亚朝野对此反应平静，政局未出现任何动荡，高层权力交接逐步平稳过渡。
改善服务门诊一站式服务 平台实施方案 Prepared on 22 November 2020
优化门诊服务实施方案一、目的： 优化服务流程，切实改善患者就医感受，缩短排队等候时间，和谐医患关系，不断提升医患满意度。 二、实施方向： 成立门诊服务中心“一站式服务平台”。 三、具体方案如下： 1．组织管理：医院一站式服务接待中心由门诊办公室统一管理，负责人员培训；护理部协助人员调配；计算机中心协助信息化维护；后勤部协助服务台建设。 2．服务项目：预检分诊、传染病筛查、预约挂号（取号）、健康咨询、投诉接待、发放检验报告、提供便民服务，接待双向转诊上转患者。 3．服务时间：上午7:30~下午17:00，双休日、节假日照常服务。4．接待中心设立在门诊一楼显眼位置，服务电话xx。 四、具体工作内容： 1.工作人员由门诊部按服务项目合理安排，服务到位。对不能立即 落实的，进行登记，并与相关部门联系，及时答复。 2.门诊代理主任现场值班，审核签章工作前移。 3.预检分诊，指引患者选择对应科室就诊，避免患者挂错科室，延 误患者就诊，减少患者退改号例数。
4.提供患者复诊预约服务，并协调相关科室安排就诊时间，指导患 者自助挂号。 5.为患者提供健教处方、专家简介、特色科室简介等资料；提供健 康咨询。 6.设意见箱，发放门诊满意度调查问卷，指导患者就诊结束后将问 卷投入意见箱。 7.接待门诊投诉，当日解决患者投诉问题。职能范围内不能解决的 问题转交投诉办公室。 8.指导患者使用检验报告自助打印机，为无自助能力的患者打印检 验报告。 9.拓展便民服务，为85岁以上患者提供轮椅，全程陪诊。 10.设立巡导志愿者岗位，安排导医、志愿者协助年老、行动不便患 者办理就诊手续。 11.接待新农合、社区转诊患者，安排就诊。 五、服务要求与管理 1.遵循“以病人为中心”的服务宗旨，落实服务接待中心的各项制度。 2.坚持规范服务，统一着装，挂牌上岗。 3.耐心、细致地回答病人的各种询问。 4.按要求认真做好工作记录。 5.服从工作安排，坚守岗位，不迟到早退，严禁脱岗。 6.巡导志愿者每两小时轮换一次。
知己的近义词反义词及知己的造句 本文是关于知己的近义词反义词及知己的造句，感谢您的阅读！ 知己的近义词反义词及知己的造句知己 基本解释：顾名思义是了解、理解、赏识自己的人，如"知己知彼，百战不殆";更常指懂你自己的挚友或密友，它是一生难求的朋友，友情的最高境界。正所谓:"士为知己者死"。 1.谓了解、理解、赏识、懂自己。 2.彼此相知而情谊深切的人。 【知己近义词】 亲信，好友，密友，心腹，挚友，深交，相知，知交，知友，知心，知音，石友，老友，至友 【知己反义词】 仇人敌人陌路 【知己造句】 1、我们想要被人爱、想拥有知己、想历经欢乐、想要安全感。 2、朋友本应是我们的亲密知己和支持者，但对于大多数人来说，有一些朋友比起帮助我们，更多的却是阻碍。 3、那么，为什么你就认为，随着年龄的增长，比起女人来男人们的知己和丰富的人际关系更少，因此一般容易更孤独呢？ 4、他成了我的朋友、我的知己、我的顾问。 5、无论在我当州长还是总统的时候，布鲁斯都是我的密友、顾问和知己。他这样的朋友人人需要，也是所有总统必须拥有的。
6、波兰斯基有着一段声名卓著的电影生涯，也是几乎所有电影界重要人物们的挚友和同事，他们是知己，是亲密的伙伴。 7、搜索引擎变成了可以帮追我们的忏悔室，知己，信得过的朋友。 8、这样看来，奥巴马国家安全团队中最具影响力的当属盖茨了――但他却是共和党人，他不会就五角大楼以外问题发表看法或成为总统知己。 9、我们的关系在二十年前就已经和平的结束了，但在网上，我又一次成为了他精神层面上的评论家，拉拉队，以及红颜知己。 10、这位“知己”，作为拍摄者，站在距离电视屏幕几英尺的地方对比着自己年轻版的形象。 11、父亲与儿子相互被形容为对方的政治扩音筒、知己和后援。 12、这对夫妻几乎没有什么至交或知己依然在世，而他们在后纳粹时期的德国也不可能会说出实话的。 13、她把我当作知己，于是，我便将她和情人之间的争吵了解得一清二楚。 14、有一种友谊不低于爱情；关系不属于暖昧；倾诉一直推心置腹；结局总是难成眷属；这就是知己！ 15、把你的治疗师当做是可以分享一切心事的知己。 16、莉莉安对我敞开心胸，我成了她的知己。 17、据盖洛普民意调查显示，在那些自我认同的保守党人中，尽管布什仍维持72%支持率，但他在共和党领导层中似乎很少有几位知
龙源期刊网 https://www.doczj.com/doc/4f5619375.html, 马来西亚华人的多种语言能力探析及启示 作者：牟瑛陈明辉 来源：《湖北函授大学学报》2018年第06期 [摘要]马来西亚华人普遍具有使用多种语言的能力。这种能力不是与生俱来的，而是在诸多因素的影响下而逐渐形成的一种社会现象，其中最重要的因素有国家教育政策的影响、家庭语言环境的影响以及社会环境的影响等。本文主要对这三个重要的影响因素进行分析并由此探讨马来西亚华人多语能力的养成对中国人学习英语的启示。 [关键词]多种语言能力；国家教育政策；家庭语言环境；社会环境 [中图分类号]TTl [文献标识码]A [文章编号]1671- 5918（ 2018） 06- 0120- 03 凡是到过马来西亚的人，想必都会对当地华人的语言‘能力印象深刻。笔者在马来西亚T 作了近两年，这期间也往往折服于当地华人在交谈过程中多种语言使用之间的无缝连接。对大多数当地华人而言，华语、英语、马来语是必备语种。除此之外，他们也能用广东话、福建话、潮州话、客家话，甚至是海南话等地方方言与人交流，有些人还能够听懂淡米尔语。而与华人共同生活在同一片国土上的马来西亚其他种族的人在这方面却逊色很多。占人口大多说的马来人除了会说马来语外，很多人甚至都不能用流利的英语进行交谈。当地华人是天生具有超常的语言天赋还是环境使然呢？每每与当地华人谈及此话题，他们表现出来的往往不是自豪，而是有些许无奈。究竟是什么原因造就了马来西亚华人的多种语言能力呢？ 一、国家教育政策的影响 马来西亚是一个多元种族、多种语言和多元文化的国家。在全国的30多个种族中，马来人、华人和印度人是人口最多的三个种族。其中，马来人约占67%，华人约占24%，印度人 约占7%。马来人的母语为马来语，华人的母语为华语（在马来西亚和新加坡，汉语标准语被称为华语），印度人的母语为淡米尔语。而从国家的层面上来说，马来语为国语，是马来西亚的官方语言。由于其曾为英国殖民地的历史原因，所以在马来西亚英语是各种族之间的通用语言，而华语是当地华人沟通的共同语，淡米尔语则主要用于印度裔人群之间的交流。 为了推行马来语，马来西亚政府作出了很大的努力，尤其是在教育上致力于实现“单元化”语言目标，强化马来语的地位。 在小学阶段，占主体的国民小学（简称“国小”）以马来语为教学语言，即使是在以华文为主要教学媒介语的国民型华文小学（简称“华小”），马来文也是必修课，而且学时不能低于政
学生一站式服务中心管理暂行办法 第一章总则 第一条为加强学生一站式服务中心的管理，深入推进满意工程，提高服务水平，特制定本办法。 第二条相关职能部门，应当在一站式服务中心设立办事窗口，统一集中办理学生相关事务。 第三条一站式服务中心工作人员应当热情为学生服务，努力提高办事水平，接受监督。 第四条一站式服务中心主要服务事项：学籍管理、考试认证、奖学金、助学金、助学实习、助学贷款、就业指导、技能培训、水电费查询、户籍、保险、网络服务、费用结算、心理咨询、问题投诉、意见建议等。 第五条一站式服务中心工作时间:周一到周五学校公布的正常上班时间。 第二章窗口管理 第六条一站式服务中心的服务事项，由有关职能部门负责按国家法律法规、政策和学校有关规定办理。 第七条同一个服务事项涉及三个及以上职能部门的，按照“统一受理、同步审批、限时办结”的原则，由一站式服务中心对各项流程进行整合后打包办理。 第八条有关职能部门对服务事项的设定依据、申办条件、申报材料、办理程序、收费依据及标准等内容应进行最终解释。 第三章窗口人员管理 第九条在一站式服务中心设立办事窗口的部门，应当选派素质较高，办事能力强的人员到一站式服务中心负责窗口工作。 第十条各部门派到一站式服务中心各窗口工作的人员，人事关系和所任职务不变。 第十一条窗口工作人员由一站式服务中心和所属部门双重管理，业务上接受原部门指导，工作考核由一站式服务中心负责。 第十二条窗口工作人员应当保持相对稳定，无特殊情况不得随意调换。调整窗口工作人员或者临时换人顶岗，事前应经一站式服务中心同意，并报分管校领导批准。 第四章办件制度 第十三条一站式服务中心办事服务实行首问负责、反馈、承诺责任制，整合审批制和费用收缴分离制等工作制度。 第十四条申请事项材料不齐全或者不符合规定的，办事窗口应当告知申请人需要补齐的全部材料及有效途径和方法。 第十五条申请人提交的申请材料齐全，符合办理条件的，应当给予办理。 第十六条按相关规定不能当场办结的事项，有关办事窗口应当承诺办结时限。承诺的办结时限不得超过5个工作日；需要呈报学校研究办理的事项，不得超过7个工作日；需要报学校上级部门审批的事项，一般应当在30个工作日内办结。
河南经贸职业学院一站式服务平台科研成果上报系统 使 用 手 册 河南经贸职业学院信息中心制
目录 1.系统概述 (1) 1.1一站式服务平台概述 (1) 1.2科研管理系统概述 (2) 2.系统登录 (2) 2.1一站式平台登录 (2) 2.2科研管理系统登录 (4) 3.论文上报 (5) 4.论著上报 (7) 5.科研成果上报 (8) 6.获奖科研成果上报 (9) 7.艺术作品竞赛成果上报 (11) 8.学生技能竞赛上报 (12) 9.个人科研综合查询 (13) 1.系统概述 1.1一站式服务平台概述 河南经贸职业学院一站式服务平台是河南经贸职业学院信息中心研制开发的一套管理信息系统，是河南经贸职业学院数字化校园的集成应用平台，主要包括内审管理、期末述职管理、期末考评、教学测评管理、公共文件柜、图书借阅、迎新管理、课时管理、一卡通查询等功能，以上各项功能根据每位用户的身份不同而分别授权使用。本平台的开发和应用，大幅提升
了学院各项管理工作的效率，并减少了工作中可能发生的错误，使学院的信息化水平达到了一个新的台阶。 1.2科研管理系统概述 科研管理系统是一站式服务平台的一个子系统，使学院教师能够方便地利用互联网上报、查询和管理自己的科研信息，并为学院有关部门统计全体教师的科研信息提供便利。系统目前包含论文上报、论文查询、论著上报、论著查询、科研成果上报、科研成果查询、获奖科研成果上报、获奖科研成果查询、艺术作品竞赛成果上报、艺术作品竞赛成果查询、学生技能竞赛上报、学生技能竞赛查询、个人科研综合查询等13项功能，其中普通教师的功能权限为论文、论著等科研成果的上报和个人科研综合查询，教务处可对每位教师上报的科研信息进行统一查询。 2. 系统登录 2.1一站式平台登录 在IE浏览器中输入https://www.doczj.com/doc/4f5619375.html,，按Enter键，即可进入如下系统登陆界面：
仿照的近义词反义词和造句 仿照的近义词 【仿制解释】：仿造：～品 【模仿解释】：个体自觉或不自觉地重复他人的行为的过程。是社会学习的重要形式之一。尤其在儿童方面，儿童的动作、语言、技能以及行为习惯、品质等的形成和发展都离不开模仿。可分为无意识模仿和有意识模仿、外部模仿和内部模仿等多种类型。 仿照的反义词 【独创解释】：独特的创造：～精神ㄧ～一格。 仿照造句 一、老师让我们仿照黑板上的图画一幅画。 二、仿照下面的句式，以“只有”开头，写一结构与之相似的复句。 三、仿照例句的句子，在下面两句的横线上补写相应的内容。 四、仿照例句，以“记忆”或“友情”开头，另写一句话。 五、仿照下面两个例句，用恰当的词语完成句子，要求前后语意关联。 六、仿照开头两句句式，通过联想，在后面两句横线上填上相应的词语。 七、仿照所给例句，用下面的词展开联想，给它一个精彩的解释。 八、仿照例句，以“你”开头，另写一个句子。 九、仿照下列句式，续写两个句子，使之与前文组成意义相关的.句子。 十、我们也仿照八股文章的笔法来一个“八股”,以毒攻毒,就叫做八大罪状吧。
十一、仿照例句，任选一种事物，写一个句子。 十二、仿照下面一句话的句式和修辞，以“时间”开头，接着写一个句子。 十三、仿照例句，以“热爱”开头，另写一句子。 十四、仿照下面的比喻形式，另写一组句子。要求选择新的本体和喻体，意思完整。 十五、根据语镜，仿照划线句子，接写两句，构成语意连贯的一段话。 十六、仿照下面句式，续写两个句式相同的比喻句。 十七、自选话题，仿照下面句子的形式和修辞，写一组排比句。 十八、仿照下面一句话的句式，仍以“人生”开头，接着写一句话。 十九、仿照例句的格式和修辞特点续写两个句子，使之与例句构成一组排比句。 二十、仿照例句，另写一个句子，要求能恰当地表达自己的愿望。 二十一、仿照下面一句话的句式，接着写一句话，使之与前面的内容、句式相对应，修辞方法相同。 二十二、仿照下面一句话的句式和修辞，以“思考”开头，接着写一个句子。 二十三、仿照下面例句，从ABCD四个英文字母中选取一个，以”青春”为话题，展开想象和联想，写一段运用了比喻修辞格、意蕴丰富的话，要求不少于30字。 二十四、仿照下面例句，另写一个句子。 二十五、仿照例句，另写一个句子。 二十六、下面是毕业前夕的班会上，数学老师为同学们写的一句赠言，请你仿照它的特点，以语文老师的身份为同学们也写一句。
暗示的近义词和反义词[暗示的近义词反义词和造句] 【暗示解释】：用含蓄、间接的方式使他人的心理、行为受到影响。下面小编就给大家整理暗示的近义词，反义词和造句，供大家学习参考。 暗示近义词 默示 示意 暗指 暗意暗示反义词 明说 表明 明言暗示造句 1. 他的顶级助手已经暗示那可能就在不久之后，但是避免设定具体的日期。 2. 一些观察家甚至暗示他可能会被送到了古拉格。 3. 要有积极的心理暗示，达成目标的好处不够充分，画面不够鲜明那它对你的吸引力就不够强烈，你就不易坚持下去! 4. 不要经常去试探男人，更不要以分手做为威胁，当你经常给他这种心理暗示，他的潜意识就会做好分手的打算。 5. 向读者暗示永远不必为主角担心，他们逢凶化吉，永远会吉人天相。 6. 约她一起运动，不要一下跳跃到游泳，可以从羽毛球网球开始，展示你的体魄和运动细胞，流汗的男人毫无疑问是最性感的，有意无意的裸露与靠近，向她暗示你的运动能力。 7. 正如在上一个示例所暗示的，只有在这些对象引用内存中同一个对象时，它们才是相同的。 8. 渥太华此时打出中国牌，巧妙地对美国这种行为作出警告，暗示美国如果长期这样下去的话，美国将自食其果。 9. 团长用震撼人心的嗓音喊道，这声音对他暗示欢乐，对兵团暗示森严，对前来校阅阅兵的首长暗示迎迓之意。 10. 渥太华此时打出中国牌，巧妙地对美国这种行为作出警告，暗示美国如果长期这样下去的话，美国将自食其恶果。 11. 我们需要邀请用户，为他们描述服务产品有多少好，给他们解释为什么他们需要填那些表单并且暗示他们会因此得到利益的回报。 12. 她对暗示她在说谎的言论嗤之以鼻。 14. 在力士参孙中，整首诗都强烈暗示着弥尔顿渴望他自己也能像参孙一样，以生命为代价，与敌人同归于尽。 15. 戴维既然问将军是否看见天花板上的钉子，这就暗示着他自己已看见了，当将军做了肯定的答复后，他又说自己看不见，这显然是自相矛盾。 16. 自我暗示在我们的生活中扮演着重要角色。如果我们不加以觉察，通常它会与我们作对。相反地，如果你运用它，它的力量就可以任你使唤。 17. 纳什建议太阳招兵买马，暗示去留取决阵容实力。 18. 同时，还暗示了菊既不同流俗，就只能在此清幽高洁而又迷漾暗淡之境中任芳姿憔悴。 19. 学习哲学的最佳途径就是将它当成一个侦探故事来处理：跟踪它的每一点蛛丝马迹每一条线索与暗示，以便查出谁是真凶，谁是英雄。 20. 不要经常去试探你的伴侣，更不要以分手做为威胁，试探本身就是一种不信任，当
浅谈马来西亚的多民族政治的发展 摘要：马来西亚是一个多民族国家，主要由马来人、华人、印度人三大种族构成，他们分别占全国总人口的47%、 34%、 9%【郭继光: 《发展过程中的利益协调———印尼与马来西亚的比较分析》 , 《东南亚研究》 , 2000年第5/6期, 第88页。】他们在宗教信仰、经济水平、价值观念上都存在较大差距。因此马来西亚的民族政治发展模式与一般的民族政治发展模式有所不同,经过了多年实践和富有特色的名族政治发展以后，马来西亚逐渐成为了一个多民族融合国家，本文试根据其具体实行的多民族政策，以及其民族政治发展所提供的经验和启示等方面，来浅谈马来西亚实行多民族政治的发展过程。 关键字：马来西亚，民族，民族融合，民族政治 马来西亚在历史上是一个典型的种族分裂社会, 种族分工存在很大差异, 形成了非马来人经济实力远远超过马来人、而马来人在政治上却享有各种特权的现象。由于这种政治、经济的不平衡, 种族之间的矛盾与冲突时有发生。因此, 马来西亚历届政府都一直致力于纠正马来人与非马来人之间政治、经济等方面的不平衡, 力求消除种族冲突的根源, 实现种族融合, 从而建立一个和谐的马来西亚。 一、马来西亚民族问题的由来 马来西亚种族分裂状态的形成是从 19 世纪末20 世纪初开始的。随着马来西亚殖民地的开发, 需要大量的劳动力, 中国、印度和斯里兰卡的移民迅速涌向马来西亚。面对庞大的外来移民, 英国殖民政府没有引起高度重视, 不但没有采取任何合理的管理措施, 使华人、印度人等外来移民融入当地社会, 反而从自己的殖民利益出发, 采取分而治之的政策, 在移民与马来人之间不断地挑拨离间, 制造磨擦、挑起矛盾, 不给予移民应有的政治权力,只有具有各土邦国民身份的马来人才能在政府部门工作, 以此来讨好本地居民, 显示马来人特殊的政治地位。与此同时, 英国殖民政府还制订了《马来人保留地法》 , 将马来人固定在土地上。所以华人和印度人的经济地位较好, 尤其是华人的经济水平远远超过当地马来人, 而马来人经济状况在各种族中是最差的。由于职业分工不同, 形成了带有种族特征的人口分布状况: 华人主要居住于城市化的西部各州, 马来人主要居住于北部和东部的农村各州。这样, 在英国殖民政府的精心策划下, 马来亚呈现出“一个国家、三种社会” (马来人社会、华人社会、印度人社会)的现象。马来人、华人、印度人三大种族在政治、经济、文化等方面存在着极大的差异, 这就使得经济与种族特征、阶级与种族特征高度同构, 使马来西亚成为一个典型的种族分裂社会, 种族间的联系和交往极少, 人为地阻碍了各民族的自然融合过程, 给马来西亚的民族政治文明建设造成了极大的负面影响。 二、马来西亚民族政治发展的路径选择
放弃的近义词反义词和造句 下面就给大家整理放弃的近义词，反义词和造句，供大家学习参考。 放弃的近义词【废弃解释】：抛弃不用：把～的土地变成良田ㄧ旧的规章制度要一概～。 【丢弃解释】：扔掉;抛弃：虽是旧衣服，他也舍不得～。 放弃的反义词【保存解释】：使事物、性质、意义、作风等继续存在，不受损失或不发生变化：～古迹ㄧ～实力ㄧ～自己，消灭敌人。 放弃造句(1) 运动员要一分一分地拼，不能放弃任何一次取胜的机会。 (2) 我们不要放弃每一个成功的机会。 (3) 敌军招架不住，只好放弃阵地，狼狈而逃。 (4) 为了农村的教育事业，姐姐主动放弃了调回城市的机会。 (5) 都快爬到山顶了，你却要放弃，岂不功亏一篑?(6) 纵使遇到再大的困难，我们也要勇往直前，不轻言放弃。 (7) 逆境中的他始终没有放弃努力,仍满怀信心,期待着峰回路转的那一天。 (8) 听了同学的规劝,他如梦初醒,放弃了离家出走的想法。 (9) 因寡不敌众,我军放弃了阵地。 (10) 要日本帝国主义放弃侵华野心,无异于与虎谋皮。 (11) 永不言弃固然好，但有时放弃却也很美。
(12) 他这种放弃原则、瓦鸡陶犬的行径已经被揭露出来了。 (13) 适当放弃，做出斩钉截铁的决定，才能成为人生的赢家。 (14) 他委曲求全地放弃自己的主张，采纳了对方的意见。 (17) 我们要有愚公移山一样的斗志，坚持不懈，永远不放弃，去登上梦想的彼岸!(18) 只要有希望，就不能放弃。 (19) 为了大局着想，你应该委曲求全地放弃自己的看法。 (20) 既然考试迫在眉睫，我不得不放弃做运动。 (21) 即使没有人相信你，也不要放弃希望。 (22) 无论通往成功的路途有多艰辛，我都不会放弃。 (23) 在困难面前，你是选择坚持，还是选择放弃?(24) 无论前路多么的漫长，过程多么的艰辛，我都不会放弃并坚定地走下去。 (25) 你不要因为这点小事就英雄气短,放弃出国深造的机会。 (26) 像他这样野心勃勃的政客，怎么可能放弃追求权力呢?(27) 鲁迅有感于中国人民愚昧和麻木,很需要做发聋振聩的启蒙工作,于是他放弃学医,改用笔来战斗。 (28) 我们对真理的追求应该坚持不懈，锲而不舍，绝不能随便放弃自己的理想。 (29) 感情之事不比其他，像你这样期盼东食西宿，几个男友都捨不得放弃，最后必定落得一场空。 (30) 爷爷临终前的话刻骨铭心，一直激励着我努力学习，无论是遇到多大的困难险阻，我都不曾放弃。
互联网+一站式服务中心方案 1.1 系统架构和功能设计 1.1.1 系统架构 硬件环境 说明：本系统采用B/S 系统架构，数据库采用ORACLE 10G ,开发工具是 Microsoft Visual Stiudio 2008 ,在本公司开发框架平台上开发 。 因本系统采用B/S 架构，办事大厅无需安装本软件，直接通过IE 访问使用本系统，部署简单，维护便捷。
接口层为与三方系统进行信息对接，公务员应用采用专线或内网进行传输，公众应用通过互联网进行传输，互联网安全配置不在该子项罗列，详情参照总体集成。 硬件环境主要指，一窗式办事大厅基础改造所需的硬件，包括窗口叫号机，取号机，查询机，窗口展示屏等。 1.1.2功能设计 1.1. 2.1一窗式办事大厅 （1）办事大厅 通过多媒体智能引导对办事人员进行多维导引,做到为事人提供业务办理前的高效引导服务，减少办事人与窗口间的不必要沟通，提供窗口的办事效率和公众服务的满意度。 智能引导系统功能模块如下： 1)大厅简介 本模块主要包含介绍大厅概况，公众如何到达大厅，宣传图片视频，天气预报等信息。 1. 大厅介绍：主要以文字的形式对中心情况进行描述同时配合中心的宣传图片视频进行播放。 2. 交通指引：集中显示周边交通地图，公众可以查看大厅周边所有公共交通线路及地图信息。
2)大厅布局 采用3D地图效果，展示大厅各楼层、各部门办事窗口的分布情况。公众根据当前位置动态制定到达线路（多路径选择），帮助办事人快速到达自己所需办事窗口，实现大厅办事人员的快速倒流。 3)办事指南 系统与大厅事项目录动态管理集成，获取入驻大厅的所有事项的办事指南信息，
浅析马来西亚“国家文化”-F的“华人文化” 一、问题的提出 马来西亚是一个以马来人、华人、印度人三大民族为主体的多元民族国家，其中，华人作为第二大民族在该国总人口中所占的比重一直都很大，形成了“少数中的多数”。据统计，马来西亚2001年全国人口为2252.9万人，其中人口最多的是马来人，为1488.6万人，占66.07％；华人是第二大民族，有572万人，占25.39％；第三大民族是印度人，有167.1万人，占7.42％。华人早在二百多年前就飘洋过海来到马来半岛谋生，华人付出的艰苦劳作和生命代价为当地的开发与发展做出了不可磨灭的贡献。正如瑞天咸爵士所作的评价：“目前马来西亚各个州之所以会有今天，全靠中国人的精神和事业。” 二战中，马来西亚华人又与各民族并肩作战，共同抵抗日本侵略者，赢得了马来西亚的独立。马来西亚建国以后，当地华人本以为自己所做的巨大努力与牺牲能够得到公正 的回报，享有与马来人同等的权利，但事实并非如此。《马来西亚联邦宪法》明确规定，马来人享有政治、经济、文化等特权，任何人不得挑战。这严重损害了非马来人的利益，
也为后来“国家文化”政策的提出埋下了伏笔。1971年，马来西亚政府提出了建设所谓“国家文化”的三原则。 随着马来西亚国内外形势的不断缓和，马来人在经济上的地位不断提升，政府把发展经济当作第一要务来抓，马来西亚的民族关系朝着融洽的方向发展。尤其是20世纪90年代以来，马来西亚政府不再处处强调马来人特权，建设所谓“国家文化”的呼声也日渐平息，取而代之的是对“多元文化”的建设。但这是否就意味着马来西亚政府对待华人的政策发生了根本性转变，各民族实现了真正的平等呢?现在提出这些问题，许多人可能认为不合时宜，但笔者认为有必要重新审视马来西亚政府所谓的“国家文化”和“多元文化”。马来西亚应该从根本上实现各民族的平等，取消马来人的特权，对各族文化的发展给于同等的重视与扶持，否则所谓的“多元文化”只是给“国家文化”换了一个更为动听的名字而已。 在“多元文化”的建设上，笔者认为政府政策的转变至关重要。只有政府对各民族实行平等的文化政策，尊重并扶持各民族文化的共同发展、共同繁荣，在此基础上，各民族文化相互交流、自行整合，才可能形成真正的“国家文化”；同时，这样的“国家文化”会进一步促进各民族的交流与融合，从而有利于国家的建设和发展。如果政府不转变政策，还是在原来的框架内大谈“多元文化”的建设，其实质还是
体面的近义词|反义词及造句 我们还必须迅速采取行动，为实现社会包容和人人体面工作营造有利的环境。下面是小编精选整理的体面的近义词|反义词及造句，供您参考，欢迎大家阅读。 体面的近义词： 面子、合适、美观、颜面、场合、场面、排场、得体、好看、局面 体面的反义词： 难听、寒碜、邋遢、寒酸、难看 体面造句 1、我认为，帖在墙面上的证书，并不能使你成为一个体面的人。 2、他是那里唯一的看上去很体面的人;我认为他从来没有这样好看过。 3、所有的美国人都是好的，体面的人们每天都在践行着他们的责任。 4、美国人民慷慨、强大、体面，这并非因为我们信任我们自己，而是因为我们拥有超越我们自己的信念。 5、工人就是工人，无论他们来自哪里，他们都应该受到尊重和尊敬，至少因为他们从事正当的工作而获得体面的工资。 6、然而反之有些孩子可能就是多少有些体面的父母的不良种子这一概念却令人难以接受。
7、如果奥巴马能够成就此功，并且帮助一个体面的伊拉克落稳脚跟，奥巴马和民主党不仅是结束了伊拉克战争，而是积极从战争中挽救。 8、而且，等到年纪大了退休时，他们希望能得到尊重和体面的对待。 9、爸爸，您倒对这件事处理得很体面，而我想那可能是我一生中最糟糕的一个夜晚吧。 10、有一些积极的东西，低于预期的就业损失索赔和零售销售是体面的。 11、如果你努力工作，你就能有获得一份终生工作的机会，有着体面的薪水，良好的福利，偶尔还能得到晋升。 12、体面的和生产性的工作是消除贫困和建立自给自足最有效的方法之一。 13、同时，他是一个仁慈、温和、体面的人，一个充满爱的丈夫和父亲，一个忠实的朋友。 14、几周前我们刚讨论过平板电脑是如何作为一个体面且多产的电子设备，即使它没有完整的键盘，在进行输入时会稍慢些。 15、什么才是生活体面的标准? 16、我们还必须迅速采取行动，为实现社会包容和人人体面工作营造有利的环境。 17、她告诉我人们都担心是不是必须把孩子送到国外去学习，才能保证孩子们长大后至少能过上体面正派的生活。