arXiv subjects

Yan Zhang

Publications and source records attributed to Yan Zhang.

At least 631 records · Page 35Linked to original sources

arXiv2018

Automated segmentaiton and classification of arterioles and venules using Cascading Dilated Convolutional Neural Networks

The change of retinal vasculature is an early sign of many vascular and systematic diseases, such as diabetes and hypertension. Different behaviors of retinal arterioles and venules form an important metric to measure the disease severity. Therefore, an accurate classification of arterioles and venules is of great necessity. In this work, we propose a novel architecture of deep convolutional neural network for segmenting and classifying arterioles and venules on retinal fundus images. This network takes the original color fundus image as inputs and multi-class labels as outputs. We adopt the encoding-decoding structure (Unet) as the backbone network of our proposed model. To improve the classification accuracy, we develop a special encoding path that couples InceptionV4 modules and Cascading Dilated Convolutions (CDCs) on top of the backbone network. The model is thus able to extract and fuse high-level semantic features from multi-scale receptive fields. The proposed method has outperformed the previous state-of-the-art method on DRIVE dataset with an accuracy of 0.955 $\pm$ 0.002.

cs.CV↗

arXiv2018

Enhanced thermoelectric performance of twisted bilayer graphene nanoribbons junction

We investigate the electron transport and thermoelectric property of twisted bilayer graphene nanoribbon junction (TBGNRJ) in $0^o$, $21.8^o$, $38.2^o$ and $60^o$ rotation angles by first principles calculation with Landauer-Buttiker and Boltzmann theories. It is found that TBGNRJs exhibit negative differential resistance (NDR) in $21.8^o$ and $38.2^o$ rotation angles under $\pm$ 0.2 V bias voltage. More importantly, three peak ZT values of 2.0, 2.7 and 6.1 can be achieved in the $21.8^o$ rotation angle at 300K. The outstanding ZT values of TBGNRJs are interpreted as the combination of the reduced thermal conductivity and enhanced electrical conductivity at optimized angles.

cond-mat.mes-hall↗

arXiv2018

Identifying multi-scale communities in networks by asymptotic surprise

Optimizing statistical measures for community structure is one of the most popular strategies for community detection, but many of them lack the flexibility of resolution and thus are incompatible with multi-scale communities of networks. Here, we further studied a statistical measure of interest for community detection, asymptotic surprise, an asymptotic approximation of surprise. We discussed the critical behaviors of asymptotic surprise in phase transition of community partition theoretically. Then, according to the theoretical analysis, a multi-resolution method based on asymptotic surprise was introduced, which provides an alternative approach to study multi-scale networks, and an improved Louvain algorithm was proposed to optimize the asymptotic surprise more effectively. By a series of experimental tests in various networks, we validated the critical behaviors of the asymptotic surprise further and the effectiveness of the improved Louvain algorithm, displayed its ability to solve the first-type resolution limit and stronger tolerance against the second-type resolution limit, and confirmed its effectiveness of revealing multi-scale community structures in multi-scale networks.

physics.soc-ph↗

arXiv2018

Exploring Correlations in Multiple Facial Attributes through Graph Attention Network

Estimating multiple attributes from a single facial image gives comprehensive descriptions on the high level semantics of the face. It is naturally regarded as a multi-task supervised learning problem with a single deep CNN, in which lower layers are shared, and higher ones are task-dependent with the multi-branch structure. Within the traditional deep multi-task learning (DMTL) framework, this paper intends to fully exploit the correlations among different attributes by constructing a graph. The node in graph represents the feature vector from a particular branch for a given attribute, and the edge can be defined by either the prior knowledge or the similarity between two nodes in the embedding with a fully data-driven manner. We analyze that the attention mechanism actually takes effect in the latter case, and utilize the Graph Attention Layer (GAL) for exploring on the most relevant attribute feature and refining the task-dependant feature by considering other attributes. Experiments show that by mining the correlations among attributes, our method can improve the recognition accuracy on CelebA and LFWA dataset. And it also achieves competitive performance.

cs.CV↗

arXiv2018

ShareJIT: JIT Code Cache Sharing across Processes and Its Practical Implementation

Just-in-time (JIT) compilation coupled with code caching are widely used to improve performance in dynamic programming language implementations. These code caches, along with the associated profiling data for the hot code, however, consume significant amounts of memory. Furthermore, they incur extra JIT compilation time for their creation. On Android, the current standard JIT compiler and its code caches are not shared among processes---that is, the runtime system maintains a private code cache, and its associated data, for each runtime process. However, applications running on the same platform tend to share multiple libraries in common. Sharing cached code across multiple applications and multiple processes can lead to a reduction in memory use. It can directly reduce compile time. It can also reduce the cumulative amount of time spent interpreting code. All three of these effects can improve actual runtime performance. In this paper, we describe ShareJIT, a global code cache for JITs that can share code across multiple applications and multiple processes. We implemented ShareJIT in the context of the Android Runtime (ART), a widely used, state-of-the-art system. To increase sharing, our implementation constrains the amount of context that the JIT compiler can use to optimize the code. This exposes a fundamental tradeoff: increased specialization to a single process' context decreases the extent to which the compiled code can be shared. In ShareJIT, we limit some optimization to increase shareability. To evaluate the ShareJIT, we tested 8 popular Android apps in a total of 30 experiments. ShareJIT improved overall performance by 9% on average, while decreasing memory consumption by 16% on average and JIT compilation time by 37% on average.

cs.PL↗

arXiv2018

High Performance Visual Tracking with Circular and Structural Operators

In this paper, a novel circular and structural operator tracker (CSOT) is proposed for high performance visual tracking, it not only possesses the powerful discriminative capability of SOSVM but also efficiently inherits the superior computational efficiency of DCF. Based on the proposed circular and structural operators, a set of primal confidence score maps can be obtained by circular correlating feature maps with their corresponding structural correlation filters. Furthermore, an implicit interpolation is applied to convert the multi-resolution feature maps to the continuous domain and make all primal confidence score maps have the same spatial resolution. Then, we exploit an efficient ensemble post-processor based on relative entropy, which can coalesce primal confidence score maps and create an optimal confidence score map for more accurate localization. The target is localized on the peak of the optimal confidence score map. Besides, we introduce a collaborative optimization strategy to update circular and structural operators by iteratively training structural correlation filters, which significantly reduces computational complexity and improves robustness. Experimental results demonstrate that our approach achieves state-of-the-art performance in mean AUC scores of 71.5% and 69.4% on the OTB-2013 and OTB-2015 benchmarks respectively, and obtains a third-best expected average overlap (EAO) score of 29.8% on the VOT-2017 benchmark.

cs.CV↗

arXiv2018

Revisiting Single Image Depth Estimation: Toward Higher Resolution Maps with Accurate Object Boundaries

This paper considers the problem of single image depth estimation. The employment of convolutional neural networks (CNNs) has recently brought about significant advancements in the research of this problem. However, most existing methods suffer from loss of spatial resolution in the estimated depth maps; a typical symptom is distorted and blurry reconstruction of object boundaries. In this paper, toward more accurate estimation with a focus on depth maps with higher spatial resolution, we propose two improvements to existing approaches. One is about the strategy of fusing features extracted at different scales, for which we propose an improved network architecture consisting of four modules: an encoder, decoder, multi-scale feature fusion module, and refinement module. The other is about loss functions for measuring inference errors used in training. We show that three loss terms, which measure errors in depth, gradients and surface normals, respectively, contribute to improvement of accuracy in an complementary fashion. Experimental results show that these two improvements enable to attain higher accuracy than the current state-of-the-arts, which is given by finer resolution reconstruction, for example, with small objects and object boundaries.

cs.CV↗

arXiv2018

VERAM: View-Enhanced Recurrent Attention Model for 3D Shape Classification

Multi-view deep neural network is perhaps the most successful approach in 3D shape classification. However, the fusion of multi-view features based on max or average pooling lacks a view selection mechanism, limiting its application in, e.g., multi-view active object recognition by a robot. This paper presents VERAM, a recurrent attention model capable of actively selecting a sequence of views for highly accurate 3D shape classification. VERAM addresses an important issue commonly found in existing attention-based models, i.e., the unbalanced training of the subnetworks corresponding to next view estimation and shape classification. The classification subnetwork is easily overfitted while the view estimation one is usually poorly trained, leading to a suboptimal classification performance. This is surmounted by three essential view-enhancement strategies: 1) enhancing the information flow of gradient backpropagation for the view estimation subnetwork, 2) devising a highly informative reward function for the reinforcement training of view estimation and 3) formulating a novel loss function that explicitly circumvents view duplication. Taking grayscale image as input and AlexNet as CNN architecture, VERAM with 9 views achieves instance-level and class-level accuracy of 95:5% and 95:3% on ModelNet10, 93:7% and 92:1% on ModelNet40, both are the state-of-the-art performance under the same number of views.

cs.CV↗

arXiv2018

Ice-templated poly(vinylidene fluoride) ferroelectrets

Ferroelectrets are piezoelectrically-active polymer foams that can convert externally applied loads into electric charge. Existing processing routes used to create pores of the desired geometry and degree of alignment appropriate for ferroelectrets are based on complex mechanical stretching and chemical dissolution steps. As a simple, cost effective and environmentally friendly approach, freeze casting is able to produce aligned pores with almost all types of the materials, including polymers. In this work, we present the first demonstration of freeze casting to create polymeric ferroelectrets. The pore morphology, phase analysis, relative permittivity and direct piezoelectric charge coefficient (d33) of porous poly(vinylidene fluoride (PVDF) ferroelectrets with porosity volume fractions ranging from 24% to 78% were analysed. The long-range alignment of pore channels produced during directional freezing was shown to be beneficial in forming a highly polarised structure after breakdown of air in the pore channels via corona poling. This new method opens a way to create tailored pores and voids in ferroelectret materials for transducer applications related to sensors and vibration energy harvesting.

physics.app-ph↗

arXiv2018

Loop Restricted Existential Rules and First-order Rewritability for Query Answering

In ontology-based data access (OBDA), the classical database is enhanced with an ontology in the form of logical assertions generating new intensional knowledge. A powerful form of such logical assertions is the tuple-generating dependencies (TGDs), also called existential rules, where Horn rules are extended by allowing existential quantifiers to appear in the rule heads. In this paper we introduce a new language called loop restricted (LR) TGDs (existential rules), which are TGDs with certain restrictions on the loops embedded in the underlying rule set. We study the complexity of this new language. We show that the conjunctive query answering (CQA) under the LR TGDs is decid- able. In particular, we prove that this language satisfies the so-called bounded derivation-depth prop- erty (BDDP), which implies that the CQA is first-order rewritable, and its data complexity is in AC0 . We also prove that the combined complexity of the CQA is EXPTIME complete, while the language membership is PSPACE complete. Then we extend the LR TGDs language to the generalised loop restricted (GLR) TGDs language, and prove that this class of TGDs still remains to be first-order rewritable and properly contains most of other first-order rewritable TGDs classes discovered in the literature so far.

cs.AI↗

arXiv2018

Local Cyber-Physical Attack for Masking Line Outage and Topology Attack in Smart Grid

Malicious attacks in the power system can eventually result in a large-scale cascade failure if not attended on time. These attacks, which are traditionally classified into \emph{physical} and \emph{cyber attacks}, can be avoided by using the latest and advanced detection mechanisms. However, a new threat called \emph{cyber-physical attacks} which jointly target both the physical and cyber layers of the system to interfere the operations of the power grid is more malicious as compared with the traditional attacks. In this paper, we propose a new cyber-physical attack strategy where the transmission line is first physically disconnected, and then the line-outage event is masked, such that the control center is misled into detecting as an obvious line outage at a different position in the local area of the power system. Therefore, the topology information in the control center is interfered by our attack. We also propose a novel procedure for selecting vulnerable lines, and analyze the observability of our proposed framework. Our proposed method can effectively and continuously deceive the control center into detecting fake line-outage positions, and thereby increase the chance of cascade failure because the attention is given to the fake outage. The simulation results validate the efficiency of our proposed attack strategy.

math.OC↗

arXiv2018

Joint Offloading and Resource Allocation in Vehicular Edge Computing and Networks

The emergence of computation intensive on-vehicle applications poses a significant challenge to provide the required computation capacity and maintain high performance. Vehicular Edge Computing (VEC) is a new computing paradigm with a high potential to improve vehicular services by offloading computation-intensive tasks to the VEC servers. Nevertheless, as the computation resource of each VEC server is limited, offloading may not be efficient if all vehicles select the same VEC server to offload their tasks. To address this problem, in this paper, we propose offloading with resource allocation. We incorporate the communication and computation to derive the task processing delay. We formulate the problem as a system utility maximization problem, and then develop a low-complexity algorithm to jointly optimize offloading decision and resource allocation. Numerical results demonstrate the superior performance of our Joint Optimization of Selection and Computation (JOSC) algorithm compared to state of the art solutions.

cs.NI↗

arXiv2018

Preferred magnetic excitations in Sr$_{1-x}$Na$_x$Fe$_2$As$_2$

We have used inelastic neutron scattering to determine magnetic excitations in a single-crystal sample of Sr$_{1-x}$Na$_x$Fe$_2$As$_2$. The material's two magnetic phases, which differ in their orthorhombic and tetragonal lattice symmetries, share very similar strengths of magnetic interactions as seen from the high-energy excitations. At low energies, excitations polarized along the $c$ axis are suppressed in the tetragonal magnetic phase, in accordance with the associated reorientation of the ordered moments. Although excitations perpendicular to the $c$ axis are still prominent, only the weak $c$-axis response exhibits a spin resonant mode in the superconducting state. Our result suggests that $c$-axis polarized magnetic excitations are important for the formation of the superconductivity, and naturally explains why the critical temperature is suppressed in the tetragonal magnetic phase.

cond-mat.supr-con↗

arXiv2018

U-SLADS: Unsupervised Learning Approach for Dynamic Dendrite Sampling

Novel data acquisition schemes have been an emerging need for scanning microscopy based imaging techniques to reduce the time in data acquisition and to minimize probing radiation in sample exposure. Varies sparse sampling schemes have been studied and are ideally suited for such applications where the images can be reconstructed from a sparse set of measurements. Dynamic sparse sampling methods, particularly supervised learning based iterative sampling algorithms, have shown promising results for sampling pixel locations on the edges or boundaries during imaging. However, dynamic sampling for imaging skeleton-like objects such as metal dendrites remains difficult. Here, we address a new unsupervised learning approach using Hierarchical Gaussian Mixture Mod- els (HGMM) to dynamically sample metal dendrites. This technique is very useful if the users are interested in fast imaging the primary and secondary arms of metal dendrites in solidification process in materials science.

eess.IV↗

arXiv2018

Subwavelength resolution Fourier ptychography with hemispherical digital condensers

Fourier ptychography (FP) is a promising computational imaging technique that overcomes the physical space-bandwidth product (SBP) limit of a conventional microscope by applying angular diversity illuminations. However, to date, the effective imaging numerical aperture (NA) achievable with a commercial LED board is still limited to the range of 0.3 to 0.7 with a 4X0.1 NA objective due to the constraint of planar geometry with weak illumination brightness and attenuated signal-to-noise ratio (SNR). Thus the highest achievable half-pitch resolution is usually constrained between 500 to 1000 nm, which cannot fulfill some needs of high-resolution biomedical imaging applications. Although it is possible to improve the resolution by using a higher magnification objective with larger NA instead of enlarging the illumination NA, the SBP is suppressed to some extent, making the FP technique less appealing, since the reduction of field-of-view (FOV) is much larger than the improvement of resolution in this FP platform. Herein, in this paper, we initially present a subwavelength resolution Fourier ptychography (SRFP) platform with a hemispherical digital condenser to provide high-angle programmable plane-wave illuminations of 0.95NA, attaining a 4X0.1 NA objective with the final effective imaging performance of 1.05NA at a half-pitch resolution of 244 nm with a wavelength of 465 nm across a wide FOV of 14.60 mm2, corresponding to an SBP of 245 megapixels. Our work provides an essential step of FP towards high-NA imaging applications without scarfing the FOV, making it more practical and appealing.

physics.optics↗

arXiv2018

3D Shape Segmentation via Shape Fully Convolutional Networks

We desgin a novel fully convolutional network architecture for shapes, denoted by Shape Fully Convolutional Networks (SFCN). 3D shapes are represented as graph structures in the SFCN architecture, based on novel graph convolution and pooling operations, which are similar to convolution and pooling operations used on images. Meanwhile, to build our SFCN architecture in the original image segmentation fully convolutional network (FCN) architecture, we also design and implement a generating operation} with bridging function. This ensures that the convolution and pooling operation we have designed can be successfully applied in the original FCN architecture. In this paper, we also present a new shape segmentation approach based on SFCN. Furthermore, we allow more general and challenging input, such as mixed datasets of different categories of shapes} which can prove the ability of our generalisation. In our approach, SFCNs are trained triangles-to-triangles by using three low-level geometric features as input. Finally, the feature voting-based multi-label graph cuts is adopted to optimise the segmentation results obtained by SFCN prediction. The experiment results show that our method can effectively learn and predict mixed shape datasets of either similar or different characteristics, and achieve excellent segmentation results.

cs.CV↗

arXiv2018

A New Finitely Controllable Class of Tuple Generating Dependencies: The Triangularly-Guarded Class

In this paper we introduce a new class of tuple-generating dependencies (TGDs) called triangularly-guarded (TG) TGDs. We show that conjunctive query answering under this new class of TGDs is decidable since this new class of TGDs also satisfies the finite controllability (FC) property. We further show that this new class strictly contains some other decidable classes such as weak-acyclic, guarded, sticky and shy. In this sense, the class TG provides a unified representation of all these aforementioned classes of TGDs.

cs.DB↗

arXiv2018

Secure Routing with Power Optimization for Ad-hoc Networks

In this paper, we consider the problem of joint secure routing and transmit power optimization for a multi-hop ad-hoc network under the existence of randomly distributed eavesdroppers following a Poisson point process (PPP). Secrecy messages are delivered from a source to a destination through a multi-hop route connected by multiple legitimate relays in the network. Our goal is to minimize the end-to-end connection outage probability (COP) under the constraint of a secrecy outage probability (SOP) threshold, by optimizing the routing path and the transmit power of each hop jointly. We show that the globally optimal solution could be obtained by a two-step procedure where the optimal transmit power has a closed-form and the optimal routing path can be found by Dijkstra's algorithm. Then a friendly jammer with multiple antennas is applied to enhance the secrecy performance further, and the optimal transmit power of the jammer and each hop of the selected route is investigated. This problem can be solved optimally via an iterative outer polyblock approximation with one-dimension search algorithm. Furthermore, suboptimal transmit powers can be derived using the successive convex approximation (SCA) method with a lower complexity. Simulation results show the performance improvement of the proposed algorithms for both non-jamming and jamming scenarios, and also reveal a non-trivial trade-off between the numbers of hops and the transmit power of each hop for secure routing.

cs.IT↗