sentence1
stringlengths 16
446
| sentence2
stringlengths 14
436
|
|---|---|
more importantly , we prove that nbg can be used to control the ferroelectric gating by unidirectionally shifting the hysteretic ferroelectric doping in graphene . | we also show that by controlling the polarity and magnitude of nbg , the hysteretic ferroelectric doping in graphene can be shifted unidirectionally . |
since this coefficient is a short distance property , we may work directly at the critical point . | then the coefficient is a sum over the decompositions . |
previous work discovered that many machine learning models , including modern neural network architectures , are vulnerable to adversarial examples . | indeed , many recent works have demonstrated the vulnerability of deep learning to adversarial manipulation . |
the combination of positron emission tomography and computerized tomography scanners has become a standard component of diagnosis and staging in oncology . | the combination of positron emission tomography and computerized tomography scanners have become a standard component of diagnosis and staging in oncology . |
the dispersion relation is a necessary stability condition for the . | this dispersion relation is the same as that in the xxz model in the certain parameter region . |
graphene , a two dimensional lattice of carbon atoms arranged in a honeycomb structure , has attracted great attention in the last few years due to its unusual properties . | graphene -a two dimensional allotrope of carbon with excellent mechanical and electronic properties -has attracted much attention since it was first produced by direct exfoliation from graphite more than a decade ago . |
as for the kitti dataset , we apply the same split in which contains 700 training images and 697 test images . | as for the kitti dataset , we finetune on the same 700 training images and evaluate on the same 697 test images as in . |
a channel estimation technique was also developed by using hierarchical multi-resolution codebookbased arvs for hybrid systems . | in , a hybrid beamforming architecture is suggested for single user massive mimo systems where matching pursuit algorithm is utilized . |
d eep convolutional neural networks have achieved great success in many computer vision tasks , especially in visual recognition . | the success of convolutional neural networks dataset has propelled neural networks to achieve significant results in various visual recognition tasks . |
sennrich et al propose an approach to use target-side monolingual data to synthesize the bitexts . | sennrich et al propose to use back-translation to generate synthetic parallel data from monolingual data for nmt . |
the data were calibrated and reduced using the miriad package . | the xrt level 2 data were analyzed using standard xselect 15 routines from ftools . |
a stratification is a decomposition that is locally well behaved . | the totallity of strata is called a stratification of x . |
the 164 decoder neurons were always initialized with xavier weights without constraints . | the network weights were randomly initialized using the uniform glorot initialization method . |
absorption is the least sensitive in all circumstances . | when there is no absorption , there is a critical frequency . |
deep neural networks have gained popularity in recent years thanks to their achievements in many applications including computer vision , signal and image processing , speech recognition . | deep neural networks have demonstrated impressive performance in many fields of research with applications ranging from image classification , just to name a few . |
in recent years , convolutional neural networks have achieved superior performance in many visual tasks , such as object classification and detection . | in recent years , convolutional neural networks have achieved significant success in many computer vision tasks , including the super-resolution problem . |
recently natarajan et al proposed a generic unbiased loss function for binary classification with noisy labels . | recently , natarajan et al proposed a generic unbiased estimator for binary classification with noisy labels . |
hence , the evolutionary differential form , in contrast to the case of the exterior form , can not be closed . | but the evolutionary form differential can not be written similarly to that presented for exterior differential forms . |
here we utilize bert as the pretrained encoder for its superior performance in a range of natural language understanding tasks . | the recently introduced bert model exhibits strong performance on several language understanding benchmarks . |
our proposed approach is partly motivated by the idea in cyclegan for image-to-image translation in the absence of paired examples . | to this end , we leverage the seminal work on cyclegans , which successfully learns various image-to-image translation tasks with unpaired examples . |
generative adversarial networks have attracted much research interest since its introduction among many others . | generative adversarial networks are a novel method for statistical inference that have received a great deal of recent attention . |
first , the landau gauge is a fixed point of the renormalization group . | the landau gauge is the singular limit a 0 of more regular gauges , and contains a non-local effective drift force kgteff , eq . |
for every diameter d , there is a d-uniform , γ-invariant reduction datum . | if t has diameter 1 or 2 , then there is a vertex v which is the neighbour of all the remaining vertices and t is a threshold graph . |
we refer to for the differential calculus of tensor fields on a riemannian manifold . | we refer to for the differential calculus of tensor fields on riemannian manifolds . |
we reduced the data using standard procedures within the common astronomy software application . | we reduced the archival alma data using the common astronomy software applications package . |
if the grow-diag-final-and method is used , an alignment point between two unaligned words appearswe conducted about three hundred of experiments to determine the best possible translation from polish to english and the reverse . | if the growdiag-final-and method is used , an alignment point between two unaligned words appearswe conducted many experiments to determine the best possible translation method from polish to english , and vice versa . |
deep neural networks have achieved great impact to broad disciplines in academia and industry . | in recent years , the deep convolutional neural networks have made great breakthroughs in computer vision . |
the lamb shift is a result of the radiative corrections , which is reflected by the fact that the positive and negative parity states are not degenerate and alternate in the spectrum . | the lamb shift is a result of the radiative corrections ( which represent effects of quantum 10 4 . |
the electron exchange correlation potential is treated with the generalized gradient of the perdew-bruke-ernzerhof functional . | the exchange-correlation of electrons was treated within the generalized gradient approximation as implemented by perdew-berke-enzelhof . |
studies have suggested that programmers spend as much or more time reading and browsing code as actually writing it . | because of this complexity , programmers tend to spend more time reading and browsing code than actually writing it . |
figure 3 shows the detailed structure of our network , which is similar to the popular alexnet . | our basic network structure is similar to alexnet structure , as depicted in figure 5 . |
our calculations are carried out within the density functional theory as implemented in the ab initio package vasp . | we have employed the vienna ab initio simulation package 28 , 29 for most of the density functional theory based first-principles calculations . |
the asterisk denotes that this result is strictly valid only when also a time average is considered . | an asterisk denotes candidate β cep and spb stars . |
for mlc , we use ms-coco , a recognition task covering 80 object classes . | we also evaluate our mfrcns on the ms coco dataset , that contains images of 80 categories of objects . |
supergravity is the theory of local supersymmetry that automatically implies general coordinate invariance . | in supergravity there is a special problem with symmetries which does not occur in yang-mills theories , the problem of auxiliary fields . |
donahue et al proposed a long-term recurrent convolutional network model for conditionally embedding the video based on the task to be performed . | donahue et al proposed an end-to-end trainable recurrent convolutional network which processes video frames with a cnn , whose outputs are passed through a recurrent neural network . |
the cnn architecture that we used was a pretrained 50-layer residual network . | we used a pre-trained network trained on imagenet with a resnet-101 architecture . |
to thoroughly evaluate the social fingerprinting technique , we compared our detection results with those obtained by different state-of-the-art spambot detection techniques , namely the supervised one by yang et al . | to thoroughly evaluate the dna fingerprinting technique we compared our detection results with those obtained by several different stateof-the-art approaches , namely the supervised one by yang et al . |
in particular , convolutional neural networks have been applied to recognizing images with great success . | convolutional neural networks have recently achieved great success on various visual recognition tasks . |
our interpretation is that it may be because the probabilities of uncorrected classes played a smaller role in this case , which supports the argument that they encode useful information for training the student model . | we interpret this result as indicating that the probabilities of uncorrected classes may play a lesser role , which supports the argument that they encode useful information for training the student model . |
in order to do so , we use a variational autoencoder to compress the mnist latent space . | hence , to construct our latent space , we use the relatively straightforward framework of a variational autoencoder . |
a path algorithm for constrained estimation . | an algorithm for total variation minimization and applications . |
deep convolutional neural networks have achieved great success on many tasks across a variety of domains , such as vision . | deep convolutional neural networks have shown tremendous success in a variety of computer vision tasks , such as image classification . |
the size of a cluster can thus increase with time , because of aftershock diffusion or secondary aftershocks , as in the work by reasenberg . | the size of a cluster can thus increase with time , due to aftershock diffusion or due to secondary aftershocks , as in reasenberg . |
deep convolutional neural networks have successfully revolutionized various challenging tasks , eg , image classification . | in recent years , deep neural networks have been applied to many areas and have achieved huge success in different domains such as image classification . |
the database contains 16,128 images of 28 human subjects under 9 poses and 64 illumination conditions . | the extended yaleb database consists of 2,414 frontal face images of 38 subjects under various lighting conditions . |
randomized linear network coding schemes were shown to be sufficient in achieving the information theoretic max-flow , min-cut bound on network capacity . | in 2003 , li et al further showed that linear network coding is sufficient to achieve the optimal throughput in multicast networks . |
these difficulties are partially due to the noise inherent in the data , which is often hard to accurately model . | such obstacles are due to the noise inherent in the data , which is often hard to accurately model . |
we consider a wireless network model and we follow definitions given in . | we consider a wireless network and we follow definitions given in . |
however , the tunnel splitting strongly depends on the strength of the transverse local stray field . | when the distribution width of the local stray field is taken into account , the distribution of tunnel splitting becomes quite large . |
the process of creating the s-tree is based on inserting and splitting the node in the tree . | the process of creating the s-tree is based on inserting and splitting the nodes in the s-tree . |
deep neural networks have achieved great success across a broad range of domains , such as computer vision , speech processing and natural language processing . | deep convolutional neural networks have been successful in many computer vision tasks including image classification . |
the nlo qcd calculation describes the shape well . | the nlo calculation describes the shape well , but lies below the data . |
recently , end-to-end deep reinforcement learning has been shown to learn complex behaviours in virtual environments . | deep reinforcement learning has been shown to be able to master complex games , even with highdimensional input such as video games . |
witten , solutions of four-dimensional gauge theories via m-theory , nucl . | witten , phase transitions in m-theory and f-theory , nucl . |
recently , deep convolutional neural networks show promising performances in various computer vision tasks such as object classification , localization . | in particular , convolutional neural network architectures have enabled superior performance over alternative approaches in classification and pattern recognition problems in computer vision . |
graphene is a planar monolayer of carbon atoms tightly packed into a two-dimensional honeycomb lattice . | graphene is the name given to a single layer of carbon atoms densely packed into a benzene-ring structure . |
projections onto the first and third eigenvectors of the spectra of the non-normalised sample . | projections onto the second and third eigenvectors of the spectra of the non-normalised sample . |
dft calculations were performed with the seqquest package developed at sandia national laboratories within the generalized gradient approximation proposed by perdew , burke , and ernzerhof . | first-principles density functional theory calculations were performed using the vienna ab initio simulation package using the generalised gradient approximation of perdew , burke , and ernzerhof . |
the two-dimensional jacobian conjecture for free poisson algebras is equivalent to the two-dimensional jacobian conjecture for polynomial algebras in characteristic zero . | the fox derivatives on free poisson algebras are defined and it is proved that an analogue of the jacobian conjecture for two generated free poisson algebras is equivalent to the two-dimensional classical jacobian conjecture . |
for the rgb branch we use the convolutional layers of a pre-trained resnet-18 network . | for a fair comparison , we use 2048-dim image features from top-layer pooling units of the 101-layered resnet . |
we used the default configuration of these methods available in weka . | we use the smo implementation provided by the weka library . |
recall that , if a is maximally monotone , then j a is defined everywhere on h and nonexpansive . | notice that ifm is maximally monotone and strongly monotone , then zer m is a singleton , thus nonempty . |
spherical varieties a normal g-variety x is called spherical if b has an open orbit on x . | as an orbit , o is a submanifold of m o is an invariant space of ρ . |
to model the dependency structure directly , we use a child-sum tree-lstm , where each word in the input sentence corresponds to a node in the dependency tree . | we combine the word vectors v i in a sentence into a single vector using a tree-structured child-sum lstm , which allows an arbitrary number of children at any node . |
dft simulations have been performed with the quantum espresso package . | these calculations have been performed using the quantum espresso code . |
correlations of positions and momenta for the electronic variables calculated in the fundamental state of the atom-field system . | correlations of the electromagnetic field variables calculated in the fundamental state of the atom-field system . |
metzen et al use both adversarial and benign samples to train a cnn-based auxiliary network . | metzen et al and grosse et al trained a classifier to detect adversarial inputs . |
measurement-based quantum computation with the toric code states . | fault-tolerant quantum computation with high threshold in two dimensions . |
in this line of research , a myopic sensing policy is developed in for the case where the su is limited to sense only one channel at each slot . | in this line of research , a myopic sensing strategy is developed in for the case where the su is limited to sense only one channel at each slot . |
it can be found that dtw performs better in finding the points with similar geometric shapes to enhance the accuracy of distance measure . | compared with the widely-used euclidean distance , dtw performs better in extracting the points with similar geometric shapes leading to enhancing the accuracy of distance measure . |
the temperature inhibits this effect and reduces the splitting between the charge multiplets . | the increasing temperature has also the effect of reducing the splitting between the charge multiplets . |
recently , deep neural networks have widely applied in computer vision tasks , such as image classification so on . | recently , deep neural networks have demonstrated impressive results in image classification . |
for non-quadratic convex cost functions , iteratively reweighed least squares daubechies et al may be used . | for non-quadratic convex cost functions , iteratively reweighed least squares may be used . |
in this section , we compare our vdsh with state-ofthe-art supervised hashing methods , including sdh on image retrieval tasks . | we further compare performance of the rmgd against the state-of-the-art binary descriptors on the brown datasets . |
the optimal power allocation strategy for the symmetric case is to allow the user with the best channel to transmit at a time . | in arbitrary fading channels , the optimal user scheduling method that maximizes the sum throughput both in uplink is to select the user who has the largest channel gain at each time slot . |
in our full labeling system , the parameters for the five convolutional blocks are borrowed from vgg network . | the parameters of the first five convolution blocks are initialized from the vgg-16 net . |
the galilean 3-space g 3 is a cayley-klein space equipped with the projective metric of signature , given in . | the galilean space g 3 is one of the cayley-klein spaces associated with the projective metric of signature . |
error analysis for the three-dimensional model was given in , in which the linearized semi-implicit euler scheme with a linear galerkin fem was used . | error analysis for the three-dimensional model was given in , in which a linearized semi-implicit euler scheme with a linear galerkin fem was used . |
deep convolutional neural networks have made significant breakthroughs in many visual understanding tasks including image classification . | advances in deep learning have led to tremendous success in a wide variety of applications in visual and audio perception such as image classification . |
the wetting film will increase by deceasing α , as has been observed under dilute conditions . | the wetting film will increase by decreasing ν t as has been observed under dilute conditions . |
on the theory of extensions of j-isometric and j-symmetric operators . | on the extension problem for dissipative operators . |
for the algorithms , we examined persistent contrastive divergence and contrastive divergence . | to train the models , we used a variation of the contrastive divergence algorithm , the persistent contrastive divergence . |
convolutional neural networks are perhaps the most widely used technique in the deep learning class of machine learning algorithms . | in addition , deep convolutional neural networks are popular for feature-learning and supervised classification . |
we also discuss briefly the disturbing hard spectator contributions . | we also briefly discuss the divergence appeared in the hard spectator contributions . |
in the past few years , convolutional neural networks have made tremendous progress in learning image features and solving various computer vision tasks . | deep convolutional neural networks have already achieved tremendous success on a variety of computer vision tasks such as image classification among many others . |
lda is an unsupervised generative model which infers probability distributions of words within topics and models documents as a mixture of those topics . | lda is a generative statistical model where its objective is to find distinct topics in document collections . |
deep neural networks have significantly improved the performance of diverse data mining and computer vision applications . | deep convolutional neural networks have made significant breakthroughs in many visual understanding tasks including image classification . |
recently , zhou et al used global average pooling to generate a class-activation mapping , visualizing discriminative image regions and enabling the localization of detected concepts . | similarly , zhou et al addressed weakly-supervised object localization using global average pooling and extended their analysis to abstract concepts . |
due to the chaotic character of the string loop motion such a transformation of the energy from the oscillatory to the transitional mode is possible . | it is well known that due to the chaotic character of the motion of string loops such a transformation of the energy from the oscillatory to the linear mode is possible . |
the last category is to represent a video in one or multiple compact images and adopt available trained convnet architectures for finetuning . | the last category represents a video as one or multiple compact images and adopts available trained convnet architectures for finetuning . |
recently , deep learning has advanced in all areas of artificial intelligence , including image classification and speech recognition . | in recent years , deep learning has become a popular approach , revolutionising various fields , including computer vision . |
against this backdrop , the breakthrough result of dyer , frieze and kannan established a randomized polynomial-time algorithm for estimating the volume to within any desired accuracy . | breakthrough result of dyer , frieze and kannan established a randomized polynomial-time algorithm for estimating the volume to within any desired accuracy . |
for general binary-input memoryless channels , density evolution was applied to estimate subchannel reliability in terms of complexity . | for general binaryinput memoryless channels , de was applied to calculate subchannel reliabilities in terms of complexity . |
a manifold n with a given symplectic form is called a phase space . | semiclassically the phase space is the natural setting whereas the quantum approach invokes high order perturbation theory involving a chain of off-resonant virtual states . |
the term p is a crossing symmetric polynomial p 2 . | the term p is the marginal likelihood or evidence . |
first compact examples of spin -holonomy manifolds were constructed by joyce . | the first examples of closed riemannian 8-manifolds with holonomy spinwere constructed by joyce . |
we also prove that the isomorphism is provided by integration on subanalytic singular chains . | we will also show that the isomorphism is given by integration on simplices . |
in recent years neural network models have been applied to document sentiment classification . | deep learning methods have been applied to many tasks related to sentiment analysis . |
the model is trained using the adam optimization method . | the deep learning models are trained by using the adam optimizer . |