基于非线性流形学习的人脸面部运动估计

由于人脸面部运动变化分布在一个低维非线性流形中,基于线性假设的主动外观模型采用主成分分析算法描述人脸形状的变化必然带来额外的估计误差.为降低或消除这一误差,该文提出一种改进的局部线性嵌入算法构建人脸形状-纹理空间,并将其应用于主动外观模型中.实验结果表明,不仅对于面部形变不大的人脸形状,局部线性嵌入-主动外观模型拥有更...     

Photo-Induced Effect of the Electron-Spin Resonance on the Charge-Ordered State in Pr0.65Ca0.35MnO3

Electron-spin resonance (ESR) for the charge-ordered state in well-characterized Pr_0.65Ca_0.35MnO₃ exhibited significant change of both the absorption profile and the effective spin susceptibility upon injection of laser light with photon energy of 1.17 eV provided by Nd-YAG laser. The increase of the effective spin susceptibility was clearly found out thanks to the injection of photons in the temperature range 90 K–80 K, which is below the transition temperature from the antiferromagnetic charge ordered state to the canted antiferromagnetic spin alternation state, T _CAF125 K. The temperature dependence of the change of the ESR profile excludes the possibility of heating by laser light. The present result suggests that a kind of photo-induced insulator–metal transition may occur due to propagation of the delocalized carriers via probable double exchange interaction in the charge-ordering collapsed state created by the injection of photons.     

Electron Spin Resonance in Pr0.65Ca0.35MnO3

The magnetic behavior of the distorted perovskite manganese, Pr_0.65Ca_0.35MnO₃ was studied by X-band electron spin resonance for powder samples. We observed the onset of the charge-ordered state at T _CO215 K, the antiferromagnetic transition with the peak of the ESR linewidth, H _p–p at T _AF180 K and the canted antiferromagnetic transition at T _CAF125 K associated with the abrupt increase of both the effective magnetization and H _p–p . Below 90 K, the absorption intensity profile becomes weakened with decreasing temperature suggesting the existence of some kind of magnetic disorder below 90 K, which is responsible for a part of evidence of the existence of the spin-glass state as has been proposed by Yoshizawa et al., Phys. Rev. B 52, 1689 (1996).     

Studies on the subsite structure of amylases. III. Inhibition by gluconolactone of the hydrolysis of maltodextrin catalyzed by glucoamylase from Rhizopus niveus

Inhibition by gluconic acid-1 : 5-lactone (gluconolactone) and phenyl alpha-glucoside of the hydrolysis of maltodextrin catalyzed by glucoamylase [EC 3.2.1.3] from Rhizopus niveus was investigated in relation to the subsite structure of the enzyme. Inhibition by gluconolactone was of the mixed type, whereas that by phenyl alpha-glucoside was purely competitive. These inhibition types were consistent with a theoretical prediction based on the assumption that gluconolactone and phenyl alpha-glucoside bind mainly to Subsites 1 and 2, respectively. The inhibitor constant of gluconolatone was determined to be 1.5 mM, which is in good agreement with the dissociation constant estimated by difference spectrophotometry (1.5 mM) (Ohnishi, M, et al. (1975) J. Biochem, 77, 695-703).     

All-optical wavelength conversion in a GaInAsP/InP optical gate loaded with a Bragg reflector

All-optical wavelength conversion employing a GaInAsP/InP nonlinear distributed-feedback waveguide switch is investigated. The device has the advantages of simple structure, compactness, polarity-noninverted operation, and feasibility of integration with other photonic devices such as semiconductor optical amplifier and modulators. We show that the device exhibits constant conversion efficiency around -9.2 dB in a signal-wavelength range of 1530-1560 nm. Furthermore, this device can potentially be made polarization independent.     

Modeling,Sensing, and Interpretation of Viscoelastic Contact Interface

Soft robotics is important in the next generation of robots because of the rapidly increasing need for robotics in biomedical applications and the advantages of providing a soft interface for interaction with the physical environment in service robots and other applications. It is indispensable to understand the fundamental behavior of such contact interface, typically viscoelastic, in order to accurately predict the actual elastic and temporal responses of the contact and to successfully control it. Viscoelasticity is a phenomenon of time-dependent strain and/or stress in soft materials. It is therefore important to model such behavior and to study the effects of such time-dependent strain and stress on stability and behavior at the contact interface. The contribution of this paper is the introduction of a novel latency model, which is a nonlinear model with differential equations that govern viscoelastic materials. Latency model describes various features of viscoelastic materials, such as stress relaxation and strain creep. The theoretical modeling was supported by experimental results in which we found two types of relaxation. Type I relaxation is well documented in existing literature but Type II relaxation has not been elaborated previously with the physical insights provided in this paper. The proposed theory can unify both types of time-dependent relaxation responses for modeling, sensing, and interpretation of viscoelastic contact interface.     

SENTIMENT ASSESSMENT OF TEXT BY ANALYZING LINGUISTIC FEATURES AND CONTEXTUAL VALENCE ASSIGNMENT

Text is not only an important medium to describe facts and events, but also to effectively communicate information about the writer's positive or negative sentiment underlying an opinion, or to express an affective or emotional state, such as happiness, fearfulness, surpriseness, and so on. We consider sentiment assessment and emotion sensing from text as two different problems, whereby sentiment assessment is the task that we want to solve first. Thus, this article presents an approach to sentiment assessment, i.e., the recognition of negative or positive valence of a sentence. For the purpose of sentiment recognition from text, we perform semantic dependency analysis on the semantic verb frames of each sentence, and then apply a set of rules to each dependency relation to calculate the contextual valence of the whole sentence. By employing a domain-independent, rule-based approach our system is able to automatically identify sentence-level sentiment. A linguistic tool called “SenseNet” has been developed to recognize sentiments in text, and to visualize the detected sentiments. We conducted several experiments with a variety of datasets containing data from different domains. The obtained results indicate significant performance gains over existing state-of-the-art approaches.