Face Configuration Research Articles

Face alignment using a 3D deeply-initialized ensemble of regression trees

Face alignment algorithms locate a set of landmark points in images of faces taken in unrestricted situations. State-of-the-art approaches typically fail or lose accuracy in the presence of occlusions, strong deformations, large pose variations and ambiguous configurations. In this paper we present 3DDE, a robust and efficient face alignment algorithm based on a coarse-to-fine cascade of ensembles of regression trees. It is initialized by robustly fitting a 3D face model to the probability maps produced by a convolutional neural network. With this initialization we address self-occlusions and large face rotations. Further, the regressor implicitly imposes a prior face shape on the solution, addressing occlusions and ambiguous face configurations. Its coarse-to-fine structure tackles the combinatorial explosion of parts deformation. In the experiments performed, 3DDE improves the state-of-the-art in 300W, COFW, AFLW and WFLW data sets. Finally, we perform cross-dataset experiments that reveal the existence of a significant data set bias in these benchmarks.

Effect of oxidative stress on cystine transportation by xC‾ antiporter

We performed computer simulations to investigate the effect of oxidation on the extracellular cystine (CYC) uptake by the xC− antiporter. The latter is important for killing of cancer cells. Specifically, applying molecular dynamics (MD) simulations we studied the transport of CYC across xCT, i.e., the light subunit of the xC− antiporter, in charge of bidirectional transport of CYC and glutamate. We considered the outward facing (OF) configuration of xCT, and to study the effect of oxidation, we modified the Cys327 residue, located in the vicinity of the extracellular milieu, to cysteic acid (CYO327).Our computational results showed that oxidation of Cys327 results in a free energy barrier for CYC translocation, thereby blocking the access of CYC to the substrate binding site of the OF system. The formation of the energy barrier was found to be due to the conformational changes in the channel. Analysis of the MD trajectories revealed that the reorganization of the side chains of the Tyr244 and CYO327 residues play a critical role in the OF channel blocking. Indeed, the calculated distance between Tyr244 and either Cys327 or CYO327 showed a narrowing of the channel after oxidation. The obtained free energy barrier for CYC translocation was found to be 33.9kJmol−1, indicating that oxidation of Cys327, by e.g., cold atmospheric plasma, is more effective in inhibiting the xC− antiporter than in the mutation of this amino acid to Ala (yielding a barrier of 32.4kJmol−1). The inhibition of the xC− antiporter may lead to Cys starvation in some cancer cells, eventually resulting in cancer cell death.

Crystal: A scalable and fault-tolerant Archimedean-based server-centric cloud data center network architecture

The explosive increase in adopting cloud services leads to the need for accommodating millions to even billions of servers in the supporting infrastructure of data center networks. Thus, providing highly scalable and noticeably fault-tolerant interconnecting architectures with great network capacity, limited delay, and reasonable energy consumption has become of paramount importance. In this paper, Crystal, a server-centric and recursively constructed structure, is presented. Crystal is made out of building blocks inspired by vertex configuration pattern of Archimedean solids and is intrinsically fault-tolerant mainly due to the adoption of multihoming mechanism in the building block. Crystal also scales double exponentially owing to its recursively defined architecture. Furthermore, to increase the routing efficiency and manage different kinds of failures in the network, two exclusively designed routing algorithms are proposed. Theoretical analysis and simulation results all witness that Crystal is highly fault-tolerant and can provide good network capacity and low latency to support delay-sensitive and data-intensive applications. Generally, Crystal satisfies a reasonable trade-off between all the major goals of designing a cloud data center network.

Vertex configuration rules to grow an octagonal Ammann–Beenker tiling

Based on vertex configurations in the Ammann–Beenker tiling, we propose an algorithm for aggregation of square and rhombus tiles to generate an octagonal quasilattice, which mimics the growth process of a two-dimensional quasicrystal. Local matching rules with configuration selection are used to guide the way that tiles are joined to a cluster and form Ammann lines according to a generalized Fibonacci sequence. Our results reveal that vertex configuration selection can improve the performance of the algorithm, which provides an approach for growing a perfect octagonal quasiperiodic structure.

Lightweight and Effective Facial Landmark Detection using Adversarial Learning with Face Geometric Map Generative Network

Facial landmark detection plays an important role in face analysis tasks. Moreover, it is used as a prerequisite in many facial related applications, the simplicity, as well as effectiveness, is essential in the facial landmark detection. In this paper, we propose an effective facial landmark detection network and an associated learning framework with the geometric prior-generative adversarial network. The geometric prior-generative adversarial network consists of one generator and two discriminators. The generator consists of an encoder and two decoders. The encoder predicts facial landmark points. The decoders generate a facial inner and contour geometric map from predicted landmark points. Generating face geometric maps from predicted landmark points helps the predicted landmark points to represent the face geometric information, including shape and configuration. The discriminators determine that the given geometric maps are generated from actual landmark points or estimated landmark points. Our proposed network is end-to-end trainable, and only the encoder part is used simply as the facial landmark detector in the testing stage. To verify the effectiveness of the proposed method, we have conducted comprehensive experiments with benchmark data sets. The results have shown that the proposed method achieves comparable performances over recently proposed facial landmark detection methods with a simple and effective facial landmark detection network.

Transport of cystine across [formula omitted] antiporter

Extracellular cystine (CYC) uptake by xC− antiporter is important for the cell viability. Especially in cancer cells, the upregulation of xC− activity is observed, which protects these cells from intracellular oxidative stress. Hence, inhibition of the CYC uptake may eventually lead to cancer cell death. Up to now, the molecular level mechanism of the CYC uptake by xC− antiporter has not been studied in detail.In this study, we applied several different simulation techniques to investigate the transport of CYC through xCT, the light subunit of the xC− antiporter, which is responsible for the CYC and glutamate translocation. Specifically, we studied the permeation of CYC across three model systems, i.e., outward facing (OF), occluded (OCC) and inward facing (IF) configurations of xCT. We also investigated the effect of mutation of Cys327 to Ala within xCT, which was also studied experimentally in literature. This allowed us to qualitatively compare our computation results with experimental observations, and thus, to validate our simulations.In summary, our simulations provide a molecular level mechanism of the transport of CYC across the xC− antiporter, more specifically, which amino acid residues in the xC− antiporter play a key role in the uptake, transport and release of CYC.

Adjacent Vertices Can be Hard to Find by Quantum Walks

Quantum walks have been useful for designing quantum algorithms that outperform their classical versions for a variety of search problems. Most of the papers, however, consider a search space containing a single marked element. We show that if the search space contains more than one marked element, their placement may drastically affect the performance of the search. More specifically, we study search by quantum walks on general graphs and show a wide class of configurations of marked vertices, for which search by quantum walk needs Ω(N) steps, that is, it has no speed-up over the classical exhaustive search. The demonstrated configurations occur for certain placements of two or more adjacent marked vertices. The analysis is done for the two-dimensional grid and hypercube, and then is generalized for any graph. Additionally, we consider an algorithmic application of the found effect. We investigate a problem of detection of a perfect matching in a bipartite graph. We use the found effect as an algorithmic building block and construct quantum algorithm which, for a specific class of graphs, outperforms its classical analogs.

Progressive Web Application Assessment Using AHP

The development of software applications has now shifted from desktop to mobile devices. Mobile application developers have been facing hardware configuration and network communication constraints in native, hybrid, and mobile web-based software architecture for a long time. Progressive Web Application is new research in this direction, created in order to overcome the limitations of native, hybrid, and mobile web-based application architecture for mobile devices. The Progressive Web Application architecture consists of a single software application which runs on multiple operating systems (android, iOS, and window phones) efficiently, uses device memory, and is even accessible in the event of no internet connection.Many IT companies have reported overall improvement in their mobile application user base after switching to Progressive Web Application. The recent discussions about Progressive Web Application on the google platform predict a demanding future. Hence, this article seeks to investigate the validity of claims that the Progressive Web Application has the edge over native, hybrid, and mobile web application architecture. The study uses the Analytical Hierarchy Process for deciding which of the architecture types suites the development of a mobile application based on three established constraints. The study outcome will aid software development communities in comprehending the Progressive Web Application and motivate them to use it in the development of an optimized mobile application.

Quasicrystal Tilings in Three Dimensions and Their Empires

The projection method for constructing quasiperiodic tilings from a higher dimensional lattice provides a useful context for computing a quasicrystal’s vertex configurations, frequencies, and empires (forced tiles). We review the projection method within the framework of the dual relationship between the Delaunay and Voronoi cell complexes of the lattice being projected. We describe a new method for calculating empires (forced tiles) which also borrows from the dualisation formalism and which generalizes to tilings generated projections of non-cubic lattices. These techniques were used to compute the vertex configurations, frequencies and empires of icosahedral quasicrystals obtained as a projections of the D 6 and Z 6 lattices to R 3 and we present our analyses. We discuss the implications of this new generalization.

DEM Simulation of the Evolution of an Unstable Rock Face: A Modelling Procedure for Back Analysis of Rockslides

The evolution of unstable rock slopes is a discontinuous process that is typically characterised by a succession of discrete events, each one giving rise to a new configuration of the rock face. If these events are put in a wider time frame, they can be seen as a local step contributing to the overall process. The advances in recognition systems, such as laser scanning or georadar techniques, allow to build numerical models of higher and higher precision, where the topographic and geostructural configurations may be precisely reconstructed. These improved capabilities open the possibility for defining highly representative numerical models that can be used for back analysis purposes or the design of risk mitigation works. One possible drawback of such approaches is that they superimpose structural and topographic data, whose compatibility is not independent of the mechanical behaviour of the rock mass. In fact, the initial geometry depends on the (usually complex) rock slope history, which has a twofold relationship with the whole set of structural and mechanical features of the rock mass. In order to investigate this point, a series of distinct element analyses of an unstable rock face located in Bolzano province is performed. The model is characterised by a very simple geometry, and slope evolution is studied by adopting the strength reduction technique. Structural and mechanical information is obtained from an extensive in situ survey. The aim of the simulations is to show how a model based on the available geomechanical information and a minimum amount of topographic data can be used to reproduce the main topographic features of the rock slope, and to perform a back analysis of a selected case history.

Small-Scale Biological and Artificial Multidimensional Sensors for 3D Sensing.

A vast majority of existing sub-millimeter-scale sensors have a planar, 2D geometry as a result of conventional top-down lithographic procedures. However, 2D sensors often suffer from restricted sensing capability, allowing only partial measurements of 3D quantities. Here, nano/microscale sensors with different geometric (1D, 2D, and 3D) configurations are reviewed to introduce their advantages and limitations when sensing changes in quantities in 3D space. This Review categorizes sensors based on their geometric configuration and sensing capabilities. Among the sensors reviewed here, the 3D configuration sensors defined on polyhedral structures are especially advantageous when sensing spatially distributed 3D quantities. The nano- and microscale vertex configuration forming polyhedral structures enable full 3D spatial sensing due to orthogonally aligned sensing elements. Particularly, the cubic configuration leveraged in 3D sensors offers an array of diverse applications in the field of biosensing for micro-organisms and proteins, optical metamaterials for invisibility cloaking, 3D imaging, and low-power remote sensing of position and angular momentum for use in microbots. Here, various 3D sensors are compared to assess the advantages of their geometry and its impact on sensing mechanisms. 3D biosensors in nature are also explored to provide vital clues for the development of novel 3D sensors.

Strongly nonperiodic hyperbolic tilings using single vertex configuration

A strongly nonperiodic tiling is defined as a tiling that does not admit infinite cyclic symmetry. The purpose of this article is to construct, up to isomorphism, uncountably many strongly nonperiodic hyperbolic tilings with a single vertex configuration by a hyperbolic rhombus tile. We use a tile found by Margulis and Mozes [5], which admits tilings, but no tiling with a compact fundamental domain.

Dynamics of two identical mutually delay-coupled semiconductor lasers in photonic integrated circuits

We theoretically investigate a system of two mutually delay-coupled semiconductor lasers, in a face to face configuration for integration in a photonic integrated circuit. This system is described by single-mode rate equations, which are a system of delay differential equations with one fixed delay. Several bifurcation scenarios involving multistabilities are presented, followed by a comprehensive frequency analysis of the symmetric and symmetry-broken, one-color and two-color states.

Self-Similar Transformation and Vertex Configurations of the Octagonal Ammann–Beenker Tiling**Supported by the National Natural Science Foundation of China under Grant No 11674102.

Based on the matching rules for squares and rhombuses, we study the self-similar transformation and the vertex configurations of the Ammann–Beenker tiling. The structural properties of the configurations and their relations during the self-similar transformation are obtained. Our results reveal the distribution correlations of the configurations, which provide an intuitive understanding of the octagonal quasi-periodic structure and also give implications for growing perfect quasi-periodic tiling according to the local rules.

Embryonic Exposure to Valproic Acid Impairs Social Predispositions of Newly-Hatched Chicks

Biological predispositions to attend to visual cues, such as those associated with face-like stimuli or with biological motion, guide social behavior from the first moments of life and have been documented in human neonates, infant monkeys and domestic chicks. Impairments of social predispositions have been recently reported in neonates at high familial risk of Autism Spectrum Disorder (ASD). Using embryonic exposure to valproic acid (VPA), an anticonvulsant associated to increased risk of developing ASD, we modeled ASD behavioral deficits in domestic chicks. We then assessed their spontaneous social predispositions by comparing approach responses to a stimulus containing a face configuration, a stuffed hen, vs. a scrambled version of it. We found that this social predisposition was abolished in VPA-treated chicks, whereas experience-dependent mechanisms associated with filial imprinting were not affected. Our results suggest a specific effect of VPA on the development of biologically-predisposed social orienting mechanisms, opening new perspectives to investigate the neurobiological mechanisms involved in early ASD symptoms.

Crushing behaviours of folded kirigami structure with square dome shape

In this study, a new type of aluminium sandwich structure with folded square dome as core is proposed. The square dome tessellated core is folded using a single piece of aluminium sheet. Four types of folded dome structures with different base sizes and top face configurations, i.e. 10 mm closed top dome, 20 mm closed top dome, 10 mm open top dome, 20 mm open top dome are studied. A single cube strip model is numerically simulated and calibrated with the experimental results from the previous studies. Good agreement on the peak and average stress between numerical results and test data is achieved. The calibrated model is then used to simulate structural response of the proposed folded dome shape structures. The damage modes and the structural responses including average and peak stress, energy absorption, uniformity ratio and densification strain are compared among these folded structures. The proposed square dome kirigami foldcore shows good energy absorption characteristics under quasi-static loading and dynamic loading by yielding a large densification strain, a low initial peak stress and a small ratio of average stress to peak stress. In addition, unlike the existing cube strip structures, the proposed folded square dome structure shows insensitivity to the crushing speed in terms of initial peak stress and uniformity ratio. Compared with the existing tessellated kirigami foldcore of cube strip, the proposed folded square dome demonstrates a superior performance than most of Miura folded structures.

Individual differences and the effect of face configuration information in the McGurk effect

The McGurk effect, which denotes the influence of visual information on audiovisual speech perception, is less frequently observed in individuals with autism spectrum disorder (ASD) compared to those without it; the reason for this remains unclear. Several studies have suggested that facial configuration context might play a role in this difference. More specifically, people with ASD show a local processing bias for faces-that is, they process global face information to a lesser extent. This study examined the role of facial configuration context in the McGurk effect in 46 healthy students. Adopting an analogue approach using the Autism-Spectrum Quotient (AQ), we sought to determine whether this facial configuration context is crucial to previously observed reductions in the McGurk effect in people with ASD. Lip-reading and audiovisual syllable identification tasks were assessed via presentation of upright normal, inverted normal, upright Thatcher-type, and inverted Thatcher-type faces. When the Thatcher-type face was presented, perceivers were found to be sensitive to the misoriented facial characteristics, causing them to perceive a weaker McGurk effect than when the normal face was presented (this is known as the McThatcher effect). Additionally, the McGurk effect was weaker in individuals with high AQ scores than in those with low AQ scores in the incongruent audiovisual condition, regardless of their ability to read lips or process facial configuration contexts. Our findings, therefore, do not support the assumption that individuals with ASD show a weaker McGurk effect due to a difficulty in processing facial configuration context.

The role of the head in configural body processing: Behavioural and electrophysiological evidence from the inversion and scrambling effect.

The present study aimed to further explore the role of the head for configural body processing by comparing complete bodies with headless bodies and faceless heads (Experiment 1). A second aim was to further explore the role of the eye region in configural face processing (Experiment 2). Due to that, we conducted a second experiment with complete faces, eyeless faces, and eyes. In addition, we used two effects to manipulate configural processing: the effect of stimulus inversion and scrambling. The current data clearly show an inversion effect for intact bodies presented with head and faces including the eye region. Thus, the head and the eye region seem to be central for configural processes that are manipulated by the effect of stimulus inversion. Furthermore, the behavioural and electrophysiological body inversion effect depends on the intact configuration of bodies and is associated with the N170 as the face inversion effect depends on the intact face configuration. Hence, configural body processing depends not only on the presence of the head but rather on a complete representation of human bodies that includes the body and the head. Furthermore, configural face processing relies on intact and complete face representations that include faces and eyes.

Effects of geometric distortions, Gaussian blur, and contrast negation on recognition of familiar faces

ABSTRACTIt is assumed that “configural processing” is important for recognizing people whom we know. Studies have shown better discrimination of relational changes between facial features when faces are presented upright than upside down and when presented as photographic negatives. However, studies have also demonstrated tolerance for complex changes to facial configuration of familiar faces when images are geometrically distorted. In two experiments, we explored effects of linear stretching, Gaussian blur, and contrast negation on recognition of familiar or unfamiliar faces (Experiment 1) and familiar or unfamiliar company brand logos (Experiment 2). Only contrast negation severely impaired recognition of familiar faces but not unfamiliar faces or company brand logos. Effects of contrast negation were additional to linear stretching, whereas linear stretching independently had no impact on recognition. These findings highlight an important deficit in the recognition of familiar faces caused by contrast negation.

Methods for Calculating Empires in Quasicrystals

This paper reviews the empire problem for quasiperiodic tilings and the existing methods for generating the empires of the vertex configurations in quasicrystals, while introducing a new and more efficient method based on the cut-and-project technique. Using Penrose tiling as an example, this method finds the forced tiles with the restrictions in the high dimensional lattice (the mother lattice) that can be cut-and-projected into the lower dimensional quasicrystal. We compare our method to the two existing methods, namely one method that uses the algorithm of the Fibonacci chain to force the Ammann bars in order to find the forced tiles of an empire and the method that follows the work of N.G. de Bruijn on constructing a Penrose tiling as the dual to a pentagrid. This new method is not only conceptually simple and clear, but it also allows us to calculate the empires of the vertex configurations in a defected quasicrystal by reversing the configuration of the quasicrystal to its higher dimensional lattice, where we then apply the restrictions. These advantages may provide a key guiding principle for phason dynamics and an important tool for self error-correction in quasicrystal growth.