arXiv subjects

Yao Xiao

Publications and source records attributed to Yao Xiao.

At least 109 records · Page 6Linked to original sources

arXiv2020

Enhance Gender and Identity Preservation in Face Aging Simulation for Infants and Toddlers

Realistic age-progressed photos provide invaluable biometric information in a wide range of applications. In recent years, deep learning-based approaches have made remarkable progress in modeling the aging process of the human face. Nevertheless, it remains a challenging task to generate accurate age-progressed faces from infant or toddler photos. In particular, the lack of visually detectable gender characteristics and the drastic appearance changes in early life contribute to the difficulty of the task. We propose a new deep learning method inspired by the successful Conditional Adversarial Autoencoder (CAAE, 2017) model. In our approach, we extend the CAAE architecture to 1) incorporate gender information, and 2) augment the model's overall architecture with an identity-preserving component based on facial features. We trained our model using the publicly available UTKFace dataset and evaluated our model by simulating up to 100 years of aging on 1,156 male and 1,207 female infant and toddler face photos. Compared to the CAAE approach, our new model demonstrates noticeable visual improvements. Quantitatively, our model exhibits an overall gain of 77.0% (male) and 13.8% (female) in gender fidelity measured by a gender classifier for the simulated photos across the age spectrum. Our model also demonstrates a 22.4% gain in identity preservation measured by a facial recognition neural network.

cs.CV↗

arXiv2020

Batch Sequential Adaptive Designs for Global Optimization

Compared with the fixed-run designs, the sequential adaptive designs (SAD) are thought to be more efficient and effective. Efficient global optimization (EGO) is one of the most popular SAD methods for expensive black-box optimization problems. A well-recognized weakness of the original EGO in complex computer experiments is that it is serial, and hence the modern parallel computing techniques cannot be utilized to speed up the running of simulator experiments. For those multiple points EGO methods, the heavy computation and points clustering are the obstacles. In this work, a novel batch SAD method, named "accelerated EGO", is forwarded by using a refined sampling/importance resampling (SIR) method to search the points with large expected improvement (EI) values. The computation burden of the new method is much lighter, and the points clustering is also avoided. The efficiency of the proposed SAD is validated by nine classic test functions with dimension from 2 to 12. The empirical results show that the proposed algorithm indeed can parallelize original EGO, and gain much improvement compared against the other parallel EGO algorithm especially under high-dimensional case. Additionally, we also apply the new method to the hyper-parameter tuning of Support Vector Machine (SVM). Accelerated EGO obtains comparable cross validation accuracy with other methods and the CPU time can be reduced a lot due to the parallel computation and sampling method.

cs.LG↗

arXiv2020

A Vertex Cut based Framework for Load Balancing and Parallelism Optimization in Multi-core Systems

High-level applications, such as machine learning, are evolving from simple models based on multilayer perceptrons for simple image recognition to much deeper and more complex neural networks for self-driving vehicle control systems.The rapid increase in the consumption of memory and computational resources by these models demands the use of multi-core parallel systems to scale the execution of the complex emerging applications that depend on them. However, parallel programs running on high-performance computers often suffer from data communication bottlenecks, limited memory bandwidth, and synchronization overhead due to irregular critical sections. In this paper, we propose a framework to reduce the data communication and improve the scalability and performance of these applications in multi-core systems. We design a vertex cut framework for partitioning LLVM IR graphs into clusters while taking into consideration the data communication and workload balance among clusters. First, we construct LLVM graphs by compiling high-level programs into LLVM IR, instrumenting code to obtain the execution order of basic blocks and the execution time for each memory operation, and analyze data dependencies in dynamic LLVM traces. Next, we formulate the problem as Weight Balanced $p$-way Vertex Cut, and propose a generic and flexible framework, wherein four different greedy algorithms are proposed for solving this problem. Lastly, we propose a memory-centric run-time mapping of the linear time complexity to map clusters generated from the vertex cut algorithms onto a multi-core platform. We conclude that our best algorithm, WB-Libra, provides performance improvements of 1.56x and 1.86x over existing state-of-the-art approaches for 8 and 1024 clusters running on a multi-core platform, respectively.

cs.DC↗

arXiv2020

Modeling indoor-level non-pharmaceutical interventions during the COVID-19 pandemic: a pedestrian dynamics-based microscopic simulation approach

Mathematical modeling of epidemic spreading has been widely adopted to estimate the threats of epidemic diseases (i.e., the COVID-19 pandemic) as well as to evaluate epidemic control interventions. The indoor place is considered to be a significant epidemic spreading risk origin, but existing widely-used epidemic spreading models are usually limited for indoor places since the dynamic physical distance changes between people are ignored, and the empirical features of the essential and non-essential travel are not differentiated. In this paper, we introduce a pedestrian-based epidemic spreading model that is capable of modeling indoor transmission risks of diseases during people's social activities. Taking advantage of the before-and-after mobility data from the University of Maryland COVID-19 Impact Analysis Platform, it's found that people tend to spend more time in grocery stores once their travel frequencies are restricted to a low level. In other words, an increase in dwell time could balance the decrease in travel frequencies and satisfy people's demand. Based on the pedestrian-based model and the empirical evidence, combined non-pharmaceutical interventions from different operational levels are evaluated. Numerical simulations show that restrictions on people's travel frequency and open-hours of indoor places may not be universally effective in reducing average infection risks for each pedestrian who visit the place. Entry limitations can be a widely effective alternative, whereas the decision-maker needs to balance the decrease in risky contacts and the increase in queue length outside the place that may impede people from fulfilling their travel needs.

physics.soc-ph↗

arXiv2019

Anharmonic corrections to the multiphonon deep-level charge capture ab initio calculations for semiconductors

Nonradiative carrier recombination at semiconductor deep centers is of great importance to both fundamental physics and device engineering. In this letter, we provide a revised analysis of K. Huang's original nonradiative multi-phonon (NMP) theory with ab initio calculations. First, we identify at first-principle level that Huang's concise formula gives the same results as the matrix based formula, and Huang's high temperature formula provides an analytical expression for the coupling constant in Marcus theory. Secondly, the anharmonic effects are corrected by taking into account local phonon mode variation at different charge states of the defect. The corrected capture rates for defects in GaN and SiC agree well with experiments.

cond-mat.mtrl-sci↗

arXiv2019

Go left or right? Explore the side preference behavior with circle antipode experiments

Side preference is a critical behavior in conflict handling, and here the behavior is investigated with pedestrian trajectories in circle antipode experiments that own both conflicting and symmetrical participant situations. In the series of experiments, more participants(around 70\%) prefer to walk on the right side, and the statistical analyses reveal that factors such as handedness, gender, and height have no significant impacts. Further investigations show that most pedestrians actually make the side choices at the very beginning, and empirical results suggest that selecting the dominate side preference (right side in our experiments) can benefit the individual movement efficiency. To reflect the realistic side preference characteristics in simulations, a Voronoi diagram based model is introduced as well as a side preference parameter with normal distribution is formulated and calibrated. Further simulations prove that the modified model is able to reproduce realistic side preference behaviors in circle antipode experiments and other common situations.

physics.soc-ph↗

arXiv2019

Design Methodology for Energy Efficient Unmanned Aerial Vehicles

In this paper, we present a load-balancing approach to analyze and partition the UAV perception and navigation intelligence (PNI) code for parallel execution, as well as assigning each parallel computational task to a processing element in an Network-on-chip (NoC) architecture such that the total communication energy is minimized and congestion is reduced. First, we construct a data dependency graph (DDG) by converting the PNI high level program into Low Level Virtual Machine (LLVM) Intermediate Representation (IR). Second, we propose a scheduling algorithm to partition the PNI application into clusters such that (1) inter-cluster communication is minimized, (2) NoC energy is reduced and (3) the workloads of different cores are balanced for maximum parallel execution. Finally, an energy-aware mapping scheme is adopted to assign clusters onto tile-based NoCs. We validate this approach with a drone self-navigation application and the experimental results show that our optimal 32-core design achieves an average 82% energy savings and 4.7x performance speedup against the state-of-art flight controller.

cs.DC↗

arXiv2019

Template-Instance Loss for Offline Handwritten Chinese Character Recognition

The long-standing challenges for offline handwritten Chinese character recognition (HCCR) are twofold: Chinese characters can be very diverse and complicated while similarly looking, and cursive handwriting (due to increased writing speed and infrequent pen lifting) makes strokes and even characters connected together in a flowing manner. In this paper, we propose the template and instance loss functions for the relevant machine learning tasks in offline handwritten Chinese character recognition. First, the character template is designed to deal with the intrinsic similarities among Chinese characters. Second, the instance loss can reduce category variance according to classification difficulty, giving a large penalty to the outlier instance of handwritten Chinese character. Trained with the new loss functions using our deep network architecture HCCR14Layer model consisting of simple layers, our extensive experiments show that it yields state-of-the-art performance and beyond for offline HCCR.

cs.CV↗

arXiv2019

Effect of gallium doping on bubbling and helium retention in aluminum exposed to low-energy helium plasma

Surface bubbling and helium retention of pure aluminum and a solid solution of aluminum-gallium(Ga) alloy exposed to low-energy He plasma has been investigated with the fluence of 1.8E24 He/m2 at room temperature . Surface morphology observations show that after irradiation, the average size of He bubbles on Al-1.8 at.% Ga alloy is around 1.8 micro metre, smaller than that on pure Al. Ga doping noticeably increases the area density of the bubble. The thermal desorption spectroscopyshows that the release amount of He in Al-1.8 at.% Ga alloy is 1.77E21 He/m2, nearly three orders of magnitude higher than that in pure Al, whlie the He desorption peak of Al-1.8 at.% Ga alloy is 480 K, much lower than 580 K of pure Al. The results of slow positron annihilation spectroscopy (SPAS) indicate that the vacancy type defects were introduced by He plasma irradiation; and lattice distortion caused by Ga doping plays an important role in determining surface bubbling and He retention characteristics of Al-1.8 at.% Ga.

cond-mat.mtrl-sci↗

arXiv2019

Comparison of deuterium retention in tungsten exposed to deuterium plasma and gas

Deuterium(D) retention behavior in tungsten(W) exposed to deuterium plasma and gas was studied by means of thermal desorption spectroscopy (TDS): deuterium plasma exposure in which W was exposed to D plamsa with 35 eV/D at 393 K to the fluence of 3.8E24 D/m2; D2 gas charging in which W was exposed to D2 gas of 500 kPa at 773 K for 4 hours. TDS shows that the total D retention in plasma exposure W is 1.00E22 D/m2, one order of magnitude higher than that of gas charging W; however, the D2 desorption peak of gas charging W is 952 K, much higher than 691 K of plasma exposure W. The detrapping energies of deuterium were determined experimentally from the measured peak temperatures at different heating rates and were found to be 2.17 eV for gas charging W and 1.04 eV for plasma exposure W, respectively.

cond-mat.mtrl-sci↗

arXiv2019

Vehicle Detection in Deep Learning

Computer vision is developing rapidly with the support of deep learning techniques. This thesis proposes an advanced vehicle-detection model based on an improvement to classical convolutional neural networks. The advanced model was applied against a vehicle detection benchmark and was built to detect on-road objects. First, we propose a high-level architecture for our advanced model, which utilizes different state-of-the-art deep learning techniques. Then, we utilize the residual neural networks and region proposal network to achieve competitive performance according to the vehicle detection benchmark. Lastly, we describe the developing trend of vehicle detection techniques and the future direction of research.

cs.CV↗

arXiv2018

Investigation of Pedestrian Dynamics in Circle Antipode Experiments

To explore the pedestrian motion navigation and conflict reaction mechanisms in practice, we organized a series of circle antipode experiments. In the experiments, pedestrians are uniformly initialized on the circle and required to leave for their antipodal positions simultaneously. On the one hand, a conflicting area is naturally formulated in the central region due to the converged shortest routes, so the practical conflict avoidance behaviors can be fully explored. On the other hand, the symmetric experimental conditions of pedestrians, e.g., symmetric starting points, symmetric destination points, and symmetric surroundings lay the foundation for further quantitative comparisons among participants. The pedestrian trajectories in the experiments are recognized and rotated, and several aspects, e.g., the trajectory space distribution, route length, travel time, velocity distribution, and time-series, are investigated. It is found that: (1) Pedestrians prefer the right-hand side during the experiments; (2) The route length is as the law of log-normal distribution, the route potential obeys the exponential distribution, and the travel time is normally distributed as well as the speed; (3) Taking the short routes unexpectedly cost pedestrians plenty of travel time, while detour seems to be a time-saving decision. What's more, the series of experiments can be regarded as a basis of the model evaluation benefit from the serious conflicts and the symmetric conditions. The evaluation framework contains four distribution indexes and two time series indexes in space and time dimensions, and they are respectively graded A traditional social force model and a Voronoi diagram based modification are introduced to test the evaluation framework. The evaluation results show that the framework is beneficial to evaluate pedestrian models and even reflects the minor differences between the models.

physics.soc-ph↗

arXiv2018

Effect of exciton-phonon coupling on the interlayer excitons in transition metal dichalcogenides double layers

We investigate the correction of interlayer exciton binding energy in transition metal dichalcogenides double layers arising from the exciton-optical phonon coupling using the method of Lee-Low-Pines unitary transformation. We find that the binding energy varies in several tens of meV, depending on the polarizability of materials and interlayer distance between double layers. Moreover, the correction of binding energy results in the remarkable increasing of the critical temperature for the condensation of dilute excitonic gas basing on the Berezinskii-Kosterlitz-Thouless model. These results not only enrich the knowledge for the modulation of interlayer exciton, but also provide potential insights for the Bose-Einstein condensation and superfluid transport of interlayer exciton in two-dimensional heterostructures.

cond-mat.mes-hall↗

arXiv2018

Investigation of Voronoi diagram based Direction Choices Using Uni- and Bi-directional Trajectory Data

In a crowd, individuals make different motion choices such as "moving to destination", "following another pedestrian", and "making a detour". For the sake of convenience, the three direction choices are respectively called destination direction, following direction and detour direction in this paper. Here, it is found that the featured direction choices could be inspired by the shape characteristics of Voronoi diagram. To be specific, in the Voronoi cell of a pedestrian, the direction to a Voronoi node is regarded as a potential "detour" direction, and the direction perpendicular to a Voronoi link is regarded as a potential "following" direction. A pedestrian generally owns several alternative Voronoi nodes and Voronoi links in a Voronoi cell, and the optimal detour and following direction are determined by considering related factors such as deviation. Plus the destination direction which is directly pointing to the destination, the three basic direction choices are defined in a Voronoi cell. In order to evaluate the Voronoi diagram based basic directions, the empirical trajectory data in both uni- and bi-directional flow experiments are extracted. A time series method considering the step frequency is used to reduce the original trajectories' swaying phenomena which might disturb the recognition of actual forward direction. The deviations between the empirical velocity direction and the basic directions are investigated, and each velocity direction is classified into a basic direction or regarded as an inexplicable direction according to the deviations. The analysis results show that each basic direction could be a potential direction choice for a pedestrian. The combination of the three basic directions could cover most empirical velocity direction choices in both uni- and bi-directional flow experiments.

physics.soc-ph↗

arXiv2018

A Supervised STDP-based Training Algorithm for Living Neural Networks

Neural networks have shown great potential in many applications like speech recognition, drug discovery, image classification, and object detection. Neural network models are inspired by biological neural networks, but they are optimized to perform machine learning tasks on digital computers. The proposed work explores the possibilities of using living neural networks in vitro as basic computational elements for machine learning applications. A new supervised STDP-based learning algorithm is proposed in this work, which considers neuron engineering constrains. A 74.7% accuracy is achieved on the MNIST benchmark for handwritten digit recognition.

cs.NE↗

arXiv2017

Global solution to the nematic liquid crystal flows with heat effect

The temperature-dependent incompressible nematic liquid crystal flows in a bounded domain $Ω\subset\mathbb{R}^N$ ($N=2, 3$) are studied in this paper. Following Danchin's method in [J. Math. Fluid Mech., 2006], we use a localization argument to recover the maximal regularity of Stokes equation with variable viscosity, by which we first prove the local existence of strong solution, then extend it to a global one provided that the initial data is a sufficiently small perturbation around the trivial equilibrium state. This paper also generalizes Hu-Wang's result in [Commun. Math. Phys., 2010] to the non-isothermal case.

math.AP↗

arXiv2017

Correction of exciton binding energy in monolayer transition metal dichalcogenides

We theoretically investigate the corrections of exciton binding energy in monolayer transition metal dichalcogenides (TMDs) due to the exciton-optical phonon coupling in the Fr$\ddot{o}$hlich interaction model by using the linear operator combined Lee-Low-Pines variational method. We not only consider the excitons couple with the intrinsic longitudinal optical (LO) phonon modes, but also the surface optical phonon modes that induced by the polar substrates underneath the TMDs. We find that exciton binding energies are corrected in a large scale due to these exciton-optical phonon couplings. We discuss the dependences of exciton binding energy on the cut-off wave vector of optical phonon modes, the polarization parameters of materials and the interlayer distance between the polar substrates and TMDs. These results provide potential explanations for the divergence of the exciton binding energy between experiment and theory in TMDs.

cond-mat.mes-hall↗

arXiv2016

Global solution to the three-dimensional liquid crystal flows of Q-tensor model

A complex non-Newtonian fluid models the nematic liquid crystal flows confined in a bounded domain in $\mathbb{R}^3$ is considered. The system is a forced incompressible Navier-Stokes equation coupled with a parabolic type Q-tensor flows. Under suitable condition on the coefficients of the Landau-de Gennes energy functional, global strong solution for small initial data is proved.

math.AP↗