Searcharxiv⌕ Search

arXiv subjects

Jing Yu

Publications and source records attributed to Jing Yu.

At least 145 records · Page 8Linked to original sources
arXiv2019

Scene Graph Reasoning with Prior Visual Relationship for Visual Question Answering

One of the key issues of Visual Question Answering (VQA) is to reason with semantic clues in the visual content under the guidance of the question, how to model relational semantics still remains as a great challenge. To fully capture visual semantics, we propose to reason over a structured visual representation - scene graph, with embedded objects and inter-object relationships. This shows great benefit over vanilla vector representations and implicit visual relationship learning. Based on existing visual relationship models, we propose a visual relationship encoder that projects visual relationships into a learned deep semantic space constrained by visual context and language priors. Upon the constructed graph, we propose a Scene Graph Convolutional Network (SceneGCN) to jointly reason the object properties and relational semantics for the correct answer. We demonstrate the model's effectiveness and interpretability on the challenging GQA dataset and the classical VQA 2.0 dataset, remarkably achieving state-of-the-art 54.56% accuracy on GQA compared to the existing best model.

cs.MM↗
arXiv2019

Dual Reweighted Lp-Norm Minimization for Salt-and-pepper Noise Removal

The robust principal component analysis (RPCA), which aims to estimate underlying low-rank and sparse structures from the degraded observation data, has found wide applications in computer vision. It is usually replaced by the principal component pursuit (PCP) model in order to pursue the convex property, leading to the undesirable overshrink problem. In this paper, we propose a dual weighted lp-norm (DWLP) model with a more reasonable weighting rule and weaker powers, which greatly generalizes the previous work and provides a better approximation to the rank minimization problem for original matrix as well as the l0-norm minimization problem for sparse data. Moreover, an approximate closed-form solution is introduced to solve the lp-norm minimization, which has more stability in the nonconvex optimization and provides a more accurate estimation for the low-rank and sparse matrix recovery problem. We then apply the DWLP model to remove salt-and-pepper noise by exploiting the image nonlocal self-similarity. Both qualitative and quantitative experiments demonstrate that the proposed method outperforms other state-of-the-art methods. In terms of PSNR evaluation, our DWLP achieves about 7.188dB, 5.078dB, 3.854dB, 2.536dB and 0.158dB improvements over the current WSNM-RPCA under 10\% to 50\% salt-and-pepper noise with an interval 10\% respectively.

cs.CV↗
arXiv2019

Semantic Modeling of Textual Relationships in Cross-Modal Retrieval

Feature modeling of different modalities is a basic problem in current research of cross-modal information retrieval. Existing models typically project texts and images into one embedding space, in which semantically similar information will have a shorter distance. Semantic modeling of textural relationships is notoriously difficult. In this paper, we propose an approach to model texts using a featured graph by integrating multi-view textual relationships including semantic relations, statistical co-occurrence, and prior relations in the knowledge base. A dual-path neural network is adopted to learn multi-modal representations of information and cross-modal similarity measure jointly. We use a Graph Convolutional Network (GCN) for generating relation-aware text representations, and use a Convolutional Neural Network (CNN) with non-linearities for image representations. The cross-modal similarity measure is learned by distance metric learning. Experimental results show that, by leveraging the rich relational semantics in texts, our model can outperform the state-of-the-art models by 3.4% and 6.3% on accuracy on two benchmark datasets.

cs.MM↗
arXiv2018

Edge-Based Blur Kernel Estimation Using Sparse Representation and Self-Similarity

Blind image deconvolution is the problem of recovering the latent image from the only observed blurry image when the blur kernel is unknown. In this paper, we propose an edge-based blur kernel estimation method for blind motion deconvolution. In our previous work, we incorporate both sparse representation and self-similarity of image patches as priors into our blind deconvolution model to regularize the recovery of the latent image. Since almost any natural image has properties of sparsity and multi-scale self-similarity, we construct a sparsity regularizer and a cross-scale non-local regularizer based on our patch priors. It has been observed that our regularizers often favor sharp images over blurry ones only for image patches of the salient edges and thus we define an edge mask to locate salient edges that we want to apply our regularizers. Experimental results on both simulated and real blurry images demonstrate that our method outperforms existing state-of-the-art blind deblurring methods even for handling of very large blurs, thanks to the use of the edge mask.

cs.CV↗
arXiv2018

Approximability of the Eight-vertex Model

We initiate a study of the classification of approximation complexity of the eight-vertex model defined over 4-regular graphs. The eight-vertex model, together with its special case the six-vertex model, is one of the most extensively studied models in statistical physics, and can be stated as a problem of counting weighted orientations in graph theory. Our result concerns the approximability of the partition function on all 4-regular graphs, classified according to the parameters of the model. Our complexity results conform to the phase transition phenomenon from physics. We introduce a quantum decomposition of the eight-vertex model and prove a set of closure properties in various regions of the parameter space. Furthermore, we show that there are extra closure properties on 4-regular planar graphs. These regions of the parameter space are concordant with the phase transition threshold. Using these closure properties, we derive polynomial time approximation algorithms via Markov chain Monte Carlo. We also show that the eight-vertex model is NP-hard to approximate on the other side of the phase transition threshold.

cs.CC↗
arXiv2018

Modeling Text with Graph Convolutional Network for Cross-Modal Information Retrieval

Cross-modal information retrieval aims to find heterogeneous data of various modalities from a given query of one modality. The main challenge is to map different modalities into a common semantic space, in which distance between concepts in different modalities can be well modeled. For cross-modal information retrieval between images and texts, existing work mostly uses off-the-shelf Convolutional Neural Network (CNN) for image feature extraction. For texts, word-level features such as bag-of-words or word2vec are employed to build deep learning models to represent texts. Besides word-level semantics, the semantic relations between words are also informative but less explored. In this paper, we model texts by graphs using similarity measure based on word2vec. A dual-path neural network model is proposed for couple feature learning in cross-modal information retrieval. One path utilizes Graph Convolutional Network (GCN) for text modeling based on graph representations. The other path uses a neural network with layers of nonlinearities for image modeling based on off-the-shelf features. The model is trained by a pairwise similarity loss function to maximize the similarity of relevant text-image pairs and minimize the similarity of irrelevant pairs. Experimental results show that the proposed model outperforms the state-of-the-art methods significantly, with 17% improvement on accuracy for the best case.

cs.IR↗
arXiv2016

The Simulation of Non-Abelian Statistics of Majorana Fermions in Ising Chain with Z2 Symmetry

In this paper, we numerically study the non-Abelian statistics of the zero-energy Majorana fermions on the end of Majorana chain and show its application to quantum computing by mapping it to a spin model with special symmetry. In particular, by using transverse-field Ising model with Z2 symmetry, we verify the nontrivial non-Abelian statistics of Majorana fermions. Numerical evidence and comparison in both Majorana-representation and spin-representation are presented. The degenerate ground states of a symmetry protected spin chain therefore previde a promising platform for topological quantum computation.

cond-mat.mes-hall↗
arXiv2015

An effective criterion for Eulerian multizeta values in positive characteristic

Characteristic p multizeta values were initially studied by Thakur, who defined them as analogues of classical multiple zeta values of Euler. In the present paper we establish an effective criterion for Eulerian multizeta values, which characterizes when a multizeta value is a rational multiple of a power of the Carlitz period. The resulting "t-motivic" algorithm can tell whether any given multizeta value is Eulerian or not. We also prove that if zeta_A(s_1,...,s_r) is Eulerian, then zeta_A(s_2,...,s_r) has to be Eulerian. When r=2, this was conjectured (and later on conjectured for arbitrary r) by Lara Rodriguez and Thakur for the zeta-like case from numerical data. Our methods apply equally well to values of Carlitz multiple polylogarithms at algebraic points and zeta-like multizeta values.

math.NT↗
arXiv2015

A pharmacokinetic -- viral kinetic model describes the effect of alisporivir monotherapy or in combination with peg-IFN on 2 hepatitis C virologic response

Alisporivir is a cyclophilin inhibitor with demonstrated in vitro and in vivo activity against hepatitis C 11 virus (HCV). We estimated antiviral effectiveness of alisporivir alone or in combination with 12 pegylated-Inteferon (peg-IFN) in 88 patients infected with different HCV genotypes treated for four 13 weeks. The pharmacokinetics of both drugs were modeled and used as driving functions for the viral 14 kinetic model. Genotype was found to significantly affect pegylated-Inteferon effectiveness ($ε$= 86.3% 15 and 99.1% in genotype-1/4 and genotype-2/3, respectively, p\textless{}10 -7) and infected cells loss rate ($δ$= 16 0.22 vs 0.39 day -1 in genotype-1/4 and genotype-2/3, respectively, p\textless{}10 -6). Alisporivir effectiveness 17 was not significantly different across genotype and was high for doses $\ge$600 mg QD. We simulated 18 virologic responses with other alisporivir dosing regimens in HCV genotype-2/3 patients using the 19 model. Our predictions consistently matched the observed responses, demonstrating that this model 20 could be a useful tool for anticipating virologic response and optimize alisporivir-based therapies.

q-bio.QM↗
arXiv2012

Preparation of NOON State Induced by Macroscopic Quantum Tunneling in an Ising Chain

In this brief report, we propose a possible way, theoretically and experimentally, to generate a NOON state of the two degenerate ferromagnetic ground states of the Transverse Ising Model. In our scheme we employ the macroscopic quantum tunneling (MQT) effect between the two degenerate ferromagnetic ground states to realize the NOON state. Our calculation about the MQT process is based on a higher-order degenerate perturbation method. After doing a transformation, the MQT process could also be treated as the hopping of individual virtual fermions in the spin chain, which will leads to an analytical description of tunneling process. The experimental feasibility for generating the NOON state is discussed in the setup of linear ion trap.

quant-ph↗
arXiv2012

Emergent Supersymmetric Many-Body Systems in Doped Z2 Topological Spin Liquid of the Toric-Code Model

In this paper, we studied the doped Z2 topological spin liquid of the toric-code model. We found that the doped holes become supersymmetric particles. The ground state of the doped Z2 topological spin liquid becomes new matters of quantum states - supersymmetric Bose-Einstein condensation or supersymmetric superfluid. As a result, this system provides a unique example of manipulatable supersymmetric many-body system.

cond-mat.str-el↗
arXiv2012

Majorana Edge States for Z2 Topological Orders of the Wen-plaquette Model and the Toric-code Model

In this paper we study the symmetry protected Majorana edge states for the Z2 topological order of the Wen-plaquette model and the toric-code model and calculate the dispersion of the Majorana edge states. For the system with translational symmetry, the Majorana edge states are gapless and have the nodal points at k=0 and k=pi. For the edge states of the toric-code model without translational symmetry, the edge modes become gapped.

cond-mat.str-el↗
arXiv2010

A Novel Symmetry Constraint Of The Super cKdV System

A new (1+1)-dimensional integrable system, i. e. the super coupled Korteweg-de Vries (cKdV) system, has been constructed by a super extension of the well-known (1+1)-dimensional cKdV system. For this new system, a novel symmetry constraint between the potential and eigenfunction can be obtained by means of the binary nonlinearization of its Lax pairs. The constraints for even variables are explicit and the constraints for odd variables are implicit. Under the symmetry constraint, the spacial part and the temporal parts of the equations associated with the Lax pairs for the super cKdV system can be decomposed into the super finite-dimensional integrable Hamiltonian systems on the supersymmetry manifold $R^{4N|2N+2}$, whose integrals of motion are explicitly given.

nlin.SI↗
arXiv2009

The Bargmann symmetry constraint and binary nonlinearization of the super Dirac systems

An explicit Bargmann symmetry constraint is computed and its associated binary nonlinearization of Lax pairs is carried out for the super Dirac systems. Under the obtained symmetry constraint, the n-th flow of the super Dirac hierarchy is decomposed into two super finite-dimensional integrable Hamiltonian systems, defined over the supersymmetry manifold $R^{4N|2N}$ with the corresponding dynamical variables $x$ and $t_n$. The integrals of motion required for Liouville integrability are explicitly given.

nlin.SI↗
arXiv2009

Binary Nonlinearization of the Super Akns System Under an Implicit Symmetry Constraint

For the super AKNS system, an implicit symmetry constraint between the potentials and the eigenfunctions is proposed. After introducing some new variables to explicitly express potentials, the super AKNS system is decomposed into two compatible finite-dimensional super systems (x-part and $t_n$-part). Furthermore, we show that the obtained super systems are integrable super Hamiltonian systems in supersymmetry manifold $\mathbb{R}^{4N+2|2N+2}$.

nlin.SI↗
arXiv2009

Geometric Gamma values and zeta values in positive characteristic

In analogy with values of the classical Euler Gamma-function at rational numbers and the Riemann zeta-function at positive integers, we consider Thakur's geometric Gamma-function evaluated at rational arguments and Carlitz zeta-values at positive integers. We prove that, when considered together, all of the algebraic relations among these special values arise from the standard functional equations of the Gamma-function and from the Euler-Carlitz relations and Frobenius p-th power relations of the zeta-function.

math.NT↗
arXiv2009

Macroscopic Quantum Tunneling Effect of Z2 Topological Order

In this paper, macroscopic quantum tunneling (MQT) effect of Z2 topological order in the Wen-Plaquette model is studied. This kind of MQT is characterized by quantum tunneling processes of different virtual quasi-particles moving around a torus. By a high-order degenerate perturbation approach, the effective pseudo-spin models of the degenerate ground states are obtained. From these models, we get the energy splitting of the ground states, of which the results are consistent with those from exact diagonalization method

cond-mat.str-el↗