Robust face representation is imperative to highly accurate face recognition. In this work, we propose an open source face recognition method with deep representation named as VIPLFaceNet, which is a lO-layer deep con...
详细信息
Robust face representation is imperative to highly accurate face recognition. In this work, we propose an open source face recognition method with deep representation named as VIPLFaceNet, which is a lO-layer deep convolu- tional neural network with seven convolutional layers and three fully-connected layers. Compared with the well-known AlexNet, our VIPLFaceNet takes only 20% training time and 60% testing time, but achieves 40% drop in error rate on the real-world face recognition benchmark LFW. Our VIPLFaceNet achieves 98.60% mean accuracy on LFW us- ing one single network. An open-source C＋＋ SDK based on VIPLFaceNet is released under BSD license. The SDK takes about 150ms to process one face image in a single thread on an i7 desktop CPU. VIPLFaceNet provides a state-of-the-art start point for both academic and industrial face recognition applications.
The AGM postulates are for the belief revision(revision by a single belief), and the DP postulates for the iterated revision(revision by a finite sequence of beliefs). Li gave an R-calculus for R-configurations?|Γ,wh...
详细信息
The AGM postulates are for the belief revision(revision by a single belief), and the DP postulates for the iterated revision(revision by a finite sequence of beliefs). Li gave an R-calculus for R-configurations?|Γ,where ? is a set of atomic formulas or the negations of atomic formulas, and Γ is a finite set of formulas. In this paper, two deduction systems for the revision of a theory by another theory are given such that the systems are sound and complete, that is, if ?|Γ■Γ′ is provable then Γ′■ ? is consistent and Γ′- ? is a maximal subset of Γ such that(Γ′- ?) ∪ ? is consistent; and for any finite theories ? and Γ, there is a finite theoryΓ′ such that Γ′- ? is a maximal subset of Γ such that(Γ′- ?) ∪ ? is consistent, and ?|Γ ? Γ′ is ***, if?|Γ■Γ′ is provable then Γ′ satisfies the AGM and the DP postulates.
Word-embedding acts as one of the backbones of modern natural language processing(NLP).Recently,with the need for deploying NLP models to low-resource devices,there has been a surge of interest to compress word embedd...
详细信息
Word-embedding acts as one of the backbones of modern natural language processing(NLP).Recently,with the need for deploying NLP models to low-resource devices,there has been a surge of interest to compress word embeddings into hash codes or binary vectors so as to save the storage and memory ***,existing work learns to encode an embedding into a compressed representation from which the original embedding can be *** these methods aim to preserve most information of every individual word,they often fail to retain the relation between words,thus can yield large loss on certain *** this end,this paper presents Relation Reconstructive Binarization(R2B)to transform word embeddings into binary codes that can preserve the relation between *** its heart,R2B trains an auto-encoder to generate binary codes that allow reconstructing the wordby-word relations in the original embedding *** showed that our method achieved significant improvements over previous methods on a number of tasks along with a space-saving of up to 98.4%.Specifically,our method reached even better results on word similarity evaluation than the uncompressed pre-trained embeddings,and was significantly better than previous compression methods that do not consider word relations.
Most existing researches on relation extraction focus on binary flat relations like Bomln relation between a Person and a *** a large portion of objective facts de-scribed in natural language are complex,especially in...
详细信息
Most existing researches on relation extraction focus on binary flat relations like Bomln relation between a Person and a *** a large portion of objective facts de-scribed in natural language are complex,especially in professional documents in fields such as finance and biomedicine that require precise *** example,“the GDP of the United States in 2018 grew 2.9%compared with 2017”describes a growth rate relation between two other relations about the economic index,which is beyond the expressive power of binary flat ***,we propose the nested relation extraction problem and formulate it as a directed acyclic graph(DAG)structure extraction ***,we propose a solution using the Iterative Neural Network which extracts relations layer by *** proposed solution achieves 78.98 and 97.89 FI scores on two nested relation extraction tasks,namely semantic cause-and-efFect relation extraction and formula ***,we observe that nested relations are usually expressed in long sentences where entities are mentioned repetitively,which makes the annotation difficult and ***,we extend our model to incorporate a mention-insensitive mode that only requires annotations of relations on entity concepts(instead of exact mentions)while preserving most of its *** mention-insensitive model performs better than the mention sensitive model when the random level in mention selection is higher than 0.3.
In this paper,an iterative regularized super resolution (SR) algorithm considering non-Gaussian noise is *** on the assumption of a generalized Gaussian distribution for the contaminating noise,an lp norm is adopted t...
详细信息
In this paper,an iterative regularized super resolution (SR) algorithm considering non-Gaussian noise is *** on the assumption of a generalized Gaussian distribution for the contaminating noise,an lp norm is adopted to measure the data fidelity term in the cost *** the meantime,a regularization functional defined in terms of the desired high resolution (HR) image is employed,which allows for the simultaneous determination of its value and the partly reconstructed image at each iteration *** convergence is thoroughly *** results show the effectiveness of the proposed algorithm as well as its superiority to conventional SR methods.
This paper focuses on fast algorithm for computing the assignment reduct in inconsistent incomplete decision systems. It is quite inconvenient to judge the assignment reduct directly ac-cording to its definition. We p...
详细信息
This paper focuses on fast algorithm for computing the assignment reduct in inconsistent incomplete decision systems. It is quite inconvenient to judge the assignment reduct directly ac-cording to its definition. We propose the judgment theorem for the assignment reduct in the inconsistent incomplete decision system, which greatly simplifies judging this type reduct. On such basis, we derive a novel attribute significance measure and construct the fast assignment reduction algorithm （F-ARA）, intended for com-puting the assignment reduct in inconsistent incomplete decision systems. Final y, we make a comparison between F-ARA and the discernibility matrix-based method by experiments on 13 Univer-sity of California at Irvine （UCI） datasets, and the experimental results prove that F-ARA is efficient and feasible.
A staggered grid scheme is proposed to reduce both the total memory requirement and the CPU time of generating the corrected near matrix in the FFTbased methods. Two sets of Cartesian grids are used to project the sou...
详细信息
A staggered grid scheme is proposed to reduce both the total memory requirement and the CPU time of generating the corrected near matrix in the FFTbased methods. Two sets of Cartesian grids are used to project the source points and the field points, *** proposed method does not lower the efficiency of computing far matrix-vector products, compared with the traditional uniform Cartesian grid scheme. Some numerical experiments are provided to demonstrate both the correctness and the efficiency of the proposed method.
Hoare logic is a logic used as a way of specifying semantics of programming languages, which has been extended to be a separation logic to reason about mutable heap structure. In a model M of Hoare logic, each program...
详细信息
Hoare logic is a logic used as a way of specifying semantics of programming languages, which has been extended to be a separation logic to reason about mutable heap structure. In a model M of Hoare logic, each program α induces an M-computable function fα M on the universe of M; and the M-recursive functions are defined on M. It will be proved that the class of all the M-computable functions fα M induced by programs is equal to the class of all the M- recursive functions. Moreover, each M-recursive function is ∑ 1 NM -definable in M, where the universal quantifier is a num- ber quantifier ranging over the standard part of a nonstandard model M.
The analytical study of a large scale nonlinear neural network is an uneasy *** try to analyze the function of neural systems by probing into the fuzzy logical framework of the neural ceUs'dynamical *** papers inv...
详细信息
ISBN:
(纸本)0780394224
The analytical study of a large scale nonlinear neural network is an uneasy *** try to analyze the function of neural systems by probing into the fuzzy logical framework of the neural ceUs'dynamical *** papers investigate the relation between fuzzy logic and neural *** most investigations focus on finding new function of neural system by combining fuzzy logical and neural system. In this paper,a novel approach is used to understand the nonlinear dynamic characteristics of neural system by analyzing the fuzzy logic framework of neural *** is the only way to understand the behavior of a large scale nonlinear neural *** abstracting the fuzzy logical framework of a neural cell,our analysis enables the delicate design of network *** an example,a difficulty task to build a recurrent network model of primary visual cortex by common dynamical analysis can be easily completed by this kind approach.
We study the direct product decomposition of quantum many-valued algebras (QMV algebras) which generalizes the decomposition theorem of ortholattices (orthomodular lattices).In detail,for an idempo- tent element of a ...
详细信息
We study the direct product decomposition of quantum many-valued algebras (QMV algebras) which generalizes the decomposition theorem of ortholattices (orthomodular lattices).In detail,for an idempo- tent element of a given QMV algebra,if it commutes with every element of the QMV algebra,it can induce a direct product decomposition of the QMV *** the same time,we introduce the commutant C(S) of a set S in a QMV algebra,and prove that when S consists of idempotent elements,C(S) is a subalgebra of the QMV *** also generalizes the cases of orthomodular lattices.
暂无评论