arXiv subjects

Yi Fang

Publications and source records attributed to Yi Fang.

198 records · Page 11Linked to original sources

arXiv2018

Brown-York mass and positive scalar curvature II - Besse's conjecture and related problems

The Besse's conjecture was posted on the well-known book Einstein manifolds by Arthur L. Besse, which describes the critical point of Hilbert-Einstein functional with constraint of unit volume and constant scalar curvature. In this article, we show that there is an interesting connection between Besse's conjecture and positive mass theorem for Brown-York mass. With the aid of positive mass theorem, we investigate the geometric structure of the so-called CPE manifolds, which provides us further understanding of Besse's conjecture. As a related topic, we also have a discussion of corresponding results for V-static metrics.

math.DG↗

arXiv2018

MOEA/D with Angle-based Constrained Dominance Principle for Constrained Multi-objective Optimization Problems

This paper proposes a novel constraint-handling mechanism named angle-based constrained dominance principle (ACDP) embedded in a decomposition-based multi-objective evolutionary algorithm (MOEA/D) to solve constrained multi-objective optimization problems (CMOPs). To maintain the diversity of the working population, ACDP utilizes the information of the angle of solutions to adjust the dominance relation of solutions during the evolutionary process. This paper uses 14 benchmark instances to evaluate the performance of the MOEA/D with ACDP (MOEA/D-ACDP). Additionally, an engineering optimization problem (which is I-beam optimization problem) is optimized. The proposed MOEA/D-ACDP, and four other decomposition-based CMOEAs, including C-MOEA/D, MOEA/D-CDP, MOEA/D-Epsilon and MOEA/D-SR are tested by the above benchmarks and the engineering application. The experimental results manifest that MOEA/D-ACDP is significantly better than the other four CMOEAs on these test instances and the real-world case, which indicates that ACDP is more effective for solving CMOPs.

cs.NE↗

arXiv2017

Spectral conditions of complement for some graphical properties

L.H. Feng at el \cite{feng4} present sufficient conditions based on spectral radius for a graph with large minimum degree to be $s$-path-coverable and $s$-Hamiltonian. Motivated by this study, in this paper, we give the sufficient conditions for a graph with large minimum degree to be $s$-connected, $s$-edge-connected, $β$-deficient, $s$-path-coverable, $s$-Hamiltonian and $s$-edge-Hamiltonian in terms of spectral radius of its complement.

math.CO↗

arXiv2017

ArbiText: Arbitrary-Oriented Text Detection in Unconstrained Scene

Arbitrary-oriented text detection in the wild is a very challenging task, due to the aspect ratio, scale, orientation, and illumination variations. In this paper, we propose a novel method, namely Arbitrary-oriented Text (or ArbText for short) detector, for efficient text detection in unconstrained natural scene images. Specifically, we first adopt the circle anchors rather than the rectangular ones to represent bounding boxes, which is more robust to orientation variations. Subsequently, we incorporate a pyramid pooling module into the Single Shot MultiBox Detector framework, in order to simultaneously explore the local and global visual information, which can, therefore, generate more confidential detection results. Experiments on established scene-text datasets, such as the ICDAR 2015 and MSRA-TD500 datasets, have demonstrated the supe rior performance of the proposed method, compared to the state-of-the-art approaches.

cs.CV↗

arXiv2017

Spectral Radius and Hamiltonicity of graphs

In this paper, we study the Hamiltonicity of graphs with large minimum degree. Firstly, we present some conditions for a simple graph to be Hamilton-connected and traceable from every vertex in terms of the spectral radius of the graph or its complement respectively. Secondly, we give the conditions for a nearly balanced bipartite graph to be traceable in terms of spectral radius, signless Laplacian spectral radius of the graph or its quasi-complement respectively.

math.CO↗

arXiv2017

3D-A-Nets: 3D Deep Dense Descriptor for Volumetric Shapes with Adversarial Networks

Recently researchers have been shifting their focus towards learned 3D shape descriptors from hand-craft ones to better address challenging issues of the deformation and structural variation inherently present in 3D objects. 3D geometric data are often transformed to 3D Voxel grids with regular format in order to be better fed to a deep neural net architecture. However, the computational intractability of direct application of 3D convolutional nets to 3D volumetric data severely limits the efficiency (i.e. slow processing) and effectiveness (i.e. unsatisfied accuracy) in processing 3D geometric data. In this paper, powered with a novel design of adversarial networks (3D-A-Nets), we have developed a novel 3D deep dense shape descriptor (3D-DDSD) to address the challenging issues of efficient and effective 3D volumetric data processing. We developed new definition of 2D multilayer dense representation (MDR) of 3D volumetric data to extract concise but geometrically informative shape description and a novel design of adversarial networks that jointly train a set of convolution neural network (CNN), recurrent neural network (RNN) and an adversarial discriminator. More specifically, the generator network produces 3D shape features that encourages the clustering of samples from the same category with correct class label, whereas the discriminator network discourages the clustering by assigning them misleading adversarial class labels. By addressing the challenges posed by the computational inefficiency of direct application of CNN to 3D volumetric data, 3D-A-Nets can learn high-quality 3D-DSDD which demonstrates superior performance on 3D shape classification and retrieval over other state-of-the-art techniques by a great margin.

cs.CV↗

arXiv2017

Unsupervised 3D Reconstruction from a Single Image via Adversarial Learning

Recent advancements in deep learning opened new opportunities for learning a high-quality 3D model from a single 2D image given sufficient training on large-scale data sets. However, the significant imbalance between available amount of images and 3D models, and the limited availability of labeled 2D image data (i.e. manually annotated pairs between images and their corresponding 3D models), severely impacts the training of most supervised deep learning methods in practice. In this paper, driven by a novel design of adversarial networks, we have developed an unsupervised learning paradigm to reconstruct 3D models from a single 2D image, which is free of manually annotated pairwise input image and its associated 3D model. Particularly, the paradigm begins with training an adaption network via autoencoder with adversarial loss, which embeds unpaired 2D synthesized image domain with real world image domain to a shared latent vector space. Then, we jointly train a 3D deconvolutional network to transform the latent vector space to the 3D object space together with the embedding process. Our experiments verify our network's robust and superior performance in handling 3D volumetric object generation from a single 2D image.

cs.CV↗

arXiv2017

Variational Deep Semantic Hashing for Text Documents

As the amount of textual data has been rapidly increasing over the past decade, efficient similarity search methods have become a crucial component of large-scale information retrieval systems. A popular strategy is to represent original data samples by compact binary codes through hashing. A spectrum of machine learning methods have been utilized, but they often lack expressiveness and flexibility in modeling to learn effective representations. The recent advances of deep learning in a wide range of applications has demonstrated its capability to learn robust and powerful feature representations for complex data. Especially, deep generative models naturally combine the expressiveness of probabilistic generative models with the high capacity of deep neural networks, which is very suitable for text modeling. However, little work has leveraged the recent progress in deep learning for text hashing. In this paper, we propose a series of novel deep document generative models for text hashing. The first proposed model is unsupervised while the second one is supervised by utilizing document labels/tags for hashing. The third model further considers document-specific factors that affect the generation of words. The probabilistic generative formulation of the proposed models provides a principled framework for model extension, uncertainty estimation, simulation, and interpretability. Based on variational inference and reparameterization, the proposed models can be interpreted as encoder-decoder deep neural networks and thus they are capable of learning complex nonlinear distributed representations of the original documents. We conduct a comprehensive set of experiments on four public testbeds. The experimental results have demonstrated the effectiveness of the proposed supervised learning models for text hashing.

cs.IR↗

arXiv2017

An Improved Epsilon Constraint-handling Method in MOEA/D for CMOPs with Large Infeasible Regions

This paper proposes an improved epsilon constraint-handling mechanism, and combines it with a decomposition-based multi-objective evolutionary algorithm (MOEA/D) to solve constrained multi-objective optimization problems (CMOPs). The proposed constrained multi-objective evolutionary algorithm (CMOEA) is named MOEA/D-IEpsilon. It adjusts the epsilon level dynamically according to the ratio of feasible to total solutions (RFS) in the current population. In order to evaluate the performance of MOEA/D-IEpsilon, a new set of CMOPs with two and three objectives is designed, having large infeasible regions (relative to the feasible regions), and they are called LIR-CMOPs. Then the fourteen benchmarks, including LIR-CMOP1-14, are used to test MOEA/D-IEpsilon and four other decomposition-based CMOEAs, including MOEA/D-Epsilon, MOEA/D-SR, MOEA/D-CDP and C-MOEA/D. The experimental results indicate that MOEA/D-IEpsilon is significantly better than the other four CMOEAs on all of the test instances, which shows that MOEA/D-IEpsilon is more suitable for solving CMOPs with large infeasible regions. Furthermore, a real-world problem, namely the robot gripper optimization problem, is used to test the five CMOEAs. The experimental results demonstrate that MOEA/D-IEpsilon also outperforms the other four CMOEAs on this problem.

cs.NE↗

arXiv2017

A sphere theorem for Bach-flat manifolds with positive constant scalar curvature

We show a closed Bach-flat Riemannian manifold with a fixed positive constant scalar curvature has to be locally spherical if its Weyl and traceless Ricci tensors are small in the sense of either $L^\infty$ or $L^{\frac{n}{2}}$-norm. Compared with the complete non-compact case done by Kim, we apply a different method to achieve these results. These results generalize a rigidity theorem of positive Einstein manifolds due to M.-A.Singer. As an application, we can partially recover the well-known Chang-Gursky-Yang's $4$-dimensional conformal sphere theorem.

math.DG↗

arXiv2017

Complete bounded $λ$-hypersurfaces in the weighted volume-preserving mean curvature flow

In this paper, we study the complete bounded $λ$-hypersurfaces in weighted volume-preserving mean curvature flow. Firstly, we investigate the volume comparison theorem of complete bounded $λ$-hypersurfaces with $|A|\leqα$ and get some applications of the volume comparison theorem. Secondly, we consider the relation among $λ$, extrinsic radius $k$, intrinsic diameter $d$, and dimension $n$ of the complete $λ$-hypersurface, and we obtain some estimates for the intrinsic diameter and the extrinsic radius. At last, we get some topological properties of the bounded $λ$-hypersurface with some natural and general restrictions.

math.DG↗

arXiv2015

Asymptotic behavior of Palais-Smale sequences associated with fractional Yamabe type equations

In this paper, we analyze the asymptotic behavior of Palais-Smale sequences associated with fractional Yamabe type equations on an asymptotically hyperbolic Riemannian manifold. We prove that Palais-Smale sequences can be decomposed into the solution of the limit equation plus a finite number of bubbles, which are the rescaling of the fundamental solutions to the fractional Yamabe equation on Euclidean space. We also verify the non-interfering fact for multi-bubbles.

math.AP↗

arXiv2013

Performance of Joint Channel and Physical Network Coding Based on Alamouti STBC

This work considers the protograph-coded physical network coding (PNC) based on Alamouti space-time block coding (STBC) over Nakagami-fading two-way relay channels, in which both the two sources and relay possess two antennas. We first propose a novel precoding scheme at the two sources so as to implement the iterative decoder efficiently at the relay. We further address a simplified updating rule of the log-likelihood-ratio (LLR) in such a decoder. Based on the simplified LLR-updating rule and Gaussian approximation, we analyze the theoretical bit-error-rate (BER) of the system, which is shown to be consistent with the decoding thresholds and simulated results. Moreover, the theoretical analysis has lower computational complexity than the protograph extrinsic information transfer (PEXIT) algorithm. Consequently, the analysis not only provides a simple way to evaluate the error performance but also facilitates the design of the joint channel-and-PNC (JCNC) in wireless communication scenarios.

cs.IT↗

arXiv2013

Performance of a Multiple-Access DCSK-CC System over Nakagami-$m$ Fading Channels

In this paper, we propose a novel cooperative scheme to enhance the performance of multiple-access (MA) differential-chaos-shift-keying (DCSK) systems. We provide the bit-error-rate (BER) performance and throughput analyses for the new system with a decode-and-forward (DF) protocol over Nakagami-$m$ fading channels. Our simulated results not only show that this system significantly improves the BER performance as compared to the existing DCSK non-cooperative (DCSK-NC) system and the multiple-input multiple-output DCSK (MIMO-DCSK) system, but also verify the theoretical analyses. Furthermore, we show that the throughput of this system approximately equals that of the DCSK-NC system, both of which have prominent improvements over the MIMO-DCSK system. We thus believe that the proposed system can be a good framework for chaos-modulation-based wireless communications.

cs.IT↗

arXiv2013

Performance Analysis of Protograph LDPC Codes for Nakagami-$m$ Fading Relay Channels

In this paper, we investigate the error performance of the protograph (LDPC) codes over Nakagami-$m$ fading relay channels. We first calculate the decoding thresholds of the protograph codes over such channels with different fading depths (i.e., different values of $m$) by exploiting the modified protograph extrinsic information transfer (PEXIT) algorithm. Furthermore, based on the PEXIT analysis and using Gaussian approximation, we derive the bit-error-rate (BER) expressions for the error-free (EF) relaying protocol and decode-and-forward (DF) relaying protocol. We finally compare the threshold with the theoretical BER and the simulated BER results of the protograph codes. It reveals that the performance of DF protocol is approximately the same as that of EF protocol. Moreover, the theoretical BER expressions, which are shown to be reasonably consistent with the decoding thresholds and the simulated BERs, are able to evaluate the system performance and predict the decoding threshold with lower complexity as compared to the modified PEXIT algorithm. As a result, this work can facilitate the design of the protograph codes for the wireless communication systems.

cs.IT↗

arXiv2012

Performance Analysis of Protograph-based LDPC Codes with Spatial Diversity

In wireless communications, spatial diversity techniques, such as space-time block code (STBC) and single-input multiple-output (SIMO), are employed to strengthen the robustness of the transmitted signal against channel fading. This paper studies the performance of protograph-based low-density parity-check (LDPC) codes with receive antenna diversity. We first propose a modified version of the protograph extrinsic information transfer (PEXIT) algorithm and use it for deriving the threshold of the protograph codes in a single-input multiple-output (SIMO) system. We then calculate the decoding threshold and simulate the bit error rate (BER) of two protograph codes (accumulate-repeat-by-3-accumulate (AR3A) code and accumulate-repeat-by-4-jagged-accumulate (AR4JA) code), a regular (3, 6) LDPC code and two optimized irregular LDPC codes. The results reveal that the irregular codes achieve the best error performance in the low signal-to-noise-ratio (SNR) region and the AR3A code outperforms all other codes in the high-SNR region. Utilizing the theoretical analyses and the simulated results, we further discuss the effect of the diversity order on the performance of the protograph codes. Accordingly, the AR3A code stands out as a good candidate for wireless communication systems with multiple receive antennas.

cs.IT↗

arXiv2012

Performance of MIMO Relay DCSK-CD Systems over Nakagami Fading Channels

A multi-access multiple-input multiple-output (MIMO) relay differential chaos shift keying cooperative diversity (DCSK-CD) system is proposed in this paper as a comprehensive cooperation scheme, in which the relay and destination both employ multiple antennas to strengthen the robustness against signal fading in a wireless network. It is shown that, with spatial diversity gains, the bit error rate (BER) performance of the proposed system is remarkably better than the conventional DCSK non-cooperation (DCSK-NC) and DCSK cooperative communication (DCSK-CC) systems. Moreover, the exact BER and close-form expressions of the proposed system are derived over Nakagami fading channels through the moment generating function (MGF), which is shown to be highly consistent with the simulation results. Meanwhile, this paper illustrates a trade-off between the performance and the complexity, and provides a threshold for the number of relay antennas keeping the user consumed energy constant. Due to the above-mentioned advantages, the proposed system stands out as a good candidate or alternative for energy-constrained wireless communications based on chaotic modulation, especially for low-power and low-cost wireless personal area networks (WPANs).

cs.OH↗

arXiv2012

Protein Folding: The Gibbs Free Energy

The fundamental law for protein folding is the Thermodynamic Principle: the amino acid sequence of a protein determines its native structure and the native structure has the minimum Gibbs free energy. If all chemical problems can be answered by quantum mechanics, there should be a quantum mechanics derivation of Gibbs free energy formula G(X) for every possible conformation X of the protein. We apply quantum statistics to derive such a formula. For simplicity, only monomeric self folding globular proteins are covered. We point out some immediate applications of the formula. We show that the formula explains the observed phenomena very well. It gives a unified explanation to both folding and denaturation; it explains why hydrophobic effect is the driving force of protein folding and clarifies the role played by hydrogen bonding; it explains the successes and deficients of various surface area models. The formula also gives a clear kinetic force of the folding: Fi(X) = - \nablaxi G(X). This also gives a natural way to perform the ab initio prediction of protein structure, minimizing G(X) by Newton's fastest desciending method.

q-bio.BM↗