Searcharxiv⌕ Search

arXiv subjects

Fei Yang

Publications and source records attributed to Fei Yang.

138 records · Page 8Linked to original sources
arXiv2015

Quasisymmetric geometry of the Cantor circles as the Julia sets of rational maps

We give three families of parabolic rational maps and show that every Cantor set of circles as the Julia set of a non-hyperbolic rational map must be quasisymmetrically equivalent to the Julia set of one map in these families for suitable parameters. Combining a result obtained before, we give a complete classification of the Cantor circles Julia sets in the sense of quasisymmetric equivalence. Moreover, we study the regularity of the components of the Cantor circles Julia sets and establish a sufficient and necessary condition when a component of a Cantor circles Julia set is a quasicircle.

math.DS↗
arXiv2015

Executable Behaviour and the π-Calculus (extended abstract)

Reactive Turing machines extend classical Turing machines with a facility to model observable interactive behaviour. We call a behaviour executable if, and only if, it is behaviourally equivalent to the behaviour of a reactive Turing machine. In this paper, we study the relationship between executable behaviour and behaviour that can be specified in the pi-calculus. We establish that all executable behaviour can be specified in the pi-calculus up to divergence-preserving branching bisimilarity. The converse, however, is not true due to (intended) limitations of the model of reactive Turing machines. That is, the pi-calculus allows the specification of behaviour that is not executable up to divergence-preserving branching bisimilarity. Motivated by an intuitive understanding of executability, we then consider a restriction on the operational semantics of the pi-calculus that does associate with every pi-term executable behaviour, at least up to the version of branching bisimilarity that does not require the preservation of divergence.

cs.LO↗
arXiv2015

Joint time and frequency dissemination network over delay-stabilized fiber optic links

A precise fiber-based time and frequency dissemination scheme for multiple users with a tree-like branching topology is proposed. Through this scheme, ultra-stable signals can be easily accessed online anywhere along the fiber without affecting other sites. The scheme is tested through an experiment, in which a modulated frequency signal and a synchronized time signal are transferred to multiple remote sites over a delay-stabilized fiber optic links that are over 50 km long. Results show that the relative stabilities are 5E-14@1s and 2E-17@10000s. Meanwhile, compared with each site, time synchronization precision is less than 80 ps. These results can pave the way to practical applications in joint time and frequency dissemination network systems.

physics.optics↗
arXiv2014

Deep Poselets for Human Detection

We address the problem of detecting people in natural scenes using a part approach based on poselets. We propose a bootstrapping method that allows us to collect millions of weakly labeled examples for each poselet type. We use these examples to train a Convolutional Neural Net to discriminate different poselet types and separate them from the background class. We then use the trained CNN as a way to represent poselet patches with a Pose Discriminative Feature (PDF) vector -- a compact 256-dimensional feature vector that is effective at discriminating pose from appearance. We train the poselet model on top of PDF features and combine them with object-level CNNs for detection and bounding box prediction. The resulting model leads to state-of-the-art performance for human detection on the PASCAL datasets.

cs.CV↗
arXiv2013

On the dynamics of a family of renormalization transformations

We study the family of renormalization transformations of the generalized $d$--dimensional diamond hierarchical Potts model in statistical mechanic and prove that their Julia sets and non-escaping loci are always connected, where $d\geq 2$. In particular, we prove that their Julia sets can never be a Sierpiński carpet if the parameter is real. We show that the Julia set is a quasicircle if and only if the parameter lies in the unbounded capture domain of these models. Moreover, the asymptotic formula of the Hausdorff dimension of the Julia set is calculated as the parameter tends to infinity.

math.DS↗
arXiv2013

Face Recognition via Globality-Locality Preserving Projections

We present an improved Locality Preserving Projections (LPP) method, named Gloablity-Locality Preserving Projections (GLPP), to preserve both the global and local geometric structures of data. In our approach, an additional constraint of the geometry of classes is imposed to the objective function of conventional LPP for respecting some more global manifold structures. Moreover, we formulate a two-dimensional extension of GLPP (2D-GLPP) as an example to show how to extend GLPP with some other statistical techniques. We apply our works to face recognition on four popular face databases, namely ORL, Yale, FERET and LFW-A databases, and extensive experimental results demonstrate that the considered global manifold information can significantly improve the performance of LPP and the proposed face recognition methods outperform the state-of-the-arts.

cs.CV↗
arXiv2013

Rational maps whose Julia sets are Cantor circles

In this paper, we give a family of rational maps whose Julia sets are Cantor circles and show that every rational map whose Julia set is a Cantor set of circles must be topologically conjugate to one map in this family on their corresponding Julia sets. In particular, we give the specific expressions of some rational maps whose Julia sets are Cantor circles, but they are not topologically conjugate to any McMullen maps on their Julia sets. Moreover, some non-hyperbolic rational maps whose Julia sets are Cantor circles are also constructed.

math.DS↗
arXiv2013

Adaptive low rank and sparse decomposition of video using compressive sensing

We address the problem of reconstructing and analyzing surveillance videos using compressive sensing. We develop a new method that performs video reconstruction by low rank and sparse decomposition adaptively. Background subtraction becomes part of the reconstruction. In our method, a background model is used in which the background is learned adaptively as the compressive measurements are processed. The adaptive method has low latency, and is more robust than previous methods. We will present experimental results to demonstrate the advantages of the proposed method.

cs.IT↗
arXiv2013

Characteristics Analysis and State Transfer for non-Markovian Open Quantum Systems

The weak-coupled two-level open quantum system described by non-Markovian Time-convolution-less master equation is investigated in this paper. The cut-off frequency, coupling constant and transition frequency, which impact on the system's decay rate, coherence factor and purity, are investigated. The appropriate parameters used in system simulation experiments are determined by comparing analysis results of different values of parameters for the effects of system performance. The control laws used to transfer the system states are designed on the basis of the Lyapunov stability theorem. Numerical simulation experiments are implemented under the MATLAB environment. The features of the free evolution trajectory of the non-Markovian systems and the states transfer from a pure state to a desired pure state under the action of the proposed control laws are studied, respectively. By comparing the experimental results, the effectiveness of the proposed quantum Lyapunov control method applied to the state transfer in non-Markovian open quantum systems is verified. Meanwhile, the influences of different control parameters and cut-off frequencies on the system performance are analyzed.

quant-ph↗
arXiv2013

Orbit Tracking Control of Quantum Systems

The orbit tracking of free-evolutionary target system in closed quantum systems is studied in this paper. Based on the concept of system control theory, the unitary transformation is applied to change the time-dependent target function into a stationary target state so that the orbit tracking problem is changed into the state transfer one. A Lyapunov function with virtual mechanical quantity P is employed to design a control law for such a state transferring. The target states in density matrix are grouped into two classes: diagonal and non-diagonal. The specific convergent conditions for target state of diagonal mixed-states are derived. In the case that the target state is a non-diagonal superposition state, we propose a non-diagonal P construction method; if the target state is a non-diagonal mixed-state we use a unitary transformation to change it into a diagonal state and design a diagonal P. In such a way, the orbit tracking problem with arbitrary initial state is properly solved. The explicit expressions of P are derived to obtain a convergent control law. At last, the system simulation experiments are performed on a two-level quantum system and the tracking process is illustrated on the Bloch sphere.

math-ph↗
arXiv2012

Efficient and Effective Volume Visualization with Enhanced Isosurface Rendering

Compared with full volume rendering, isosurface rendering has several well recognized advantages in efficiency and accuracy. However, standard isosurface rendering has some limitations in effectiveness. First, it uses a monotone colored approach and can only visualize the geometry features of an isosurface. The lack of the capability to illustrate the material property and the internal structures behind an isosurface has been a big limitation of this method in applications. Another limitation of isosurface rendering is the difficulty to reveal physically meaningful structures, which are hidden in one or multiple isosurfaces. As such, the application requirements of extract and recombine structures of interest can not be implemented effectively with isosurface rendering. In this work, we develop an enhanced isosurface rendering technique to improve the effectiveness while maintaining the performance efficiency of the standard isosurface rendering. First, an isosurface color enhancement method is proposed to illustrate the neighborhood density and to reveal some of the internal structures. Second, we extend the structure extraction capability of isosurface rendering by enabling explicit scene exploration within a 3D-view, using surface peeling, voxel-selecting, isosurface segmentation, and multi-surface-structure visualization. Our experiments show that the color enhancement not only improves the visual fidelity of the rendering, but also reveals the internal structures without significant increase of the computational cost. Explicit scene exploration is also demonstrated as a powerful tool in some application scenarios, such as displaying multiple abdominal organs.

cs.GR↗