Extending 3D Lucas–Kanade tracking with adaptive templates for head pose estimation

The Lucas–Kanade tracker (LKT) is a commonly used method to track target objects over 2D images. The key principle behind the object tracking of an LKT is to warp the object appearance so as to minimize the difference between the warped object’s appearance and a pre-stored template. Accordingly, the 2D pose of the tracked object in terms of translation, rotation, and scaling can be recovered from the warping. To extend the LKT for 3D pose estimation, a model-based 3D LKT assumes a 3D geometric model for the target object in the 3D space and tries to infer the 3D object motion by minimizing the difference between the projected 2D image of the 3D object and the pre-stored 2D image template. In this paper, we propose an extended model-based 3D LKT for estimating 3D head poses by tracking human heads on video sequences. In contrast to the original model-based 3D LKT, which uses a template with each pixel represented by a single intensity value, the proposed model-based 3D LKT exploits an adaptive template with each template pixel modeled by a continuously updated Gaussian distribution during head tracking. This probabilistic template modeling improves the tracker’s ability to handle temporal fluctuation of pixels caused by continuous environmental changes such as varying illumination and dynamic backgrounds. Due to the new probabilistic template modeling, we reformulate the head pose estimation as a maximum likelihood estimation problem, rather than the original difference minimization procedure. Based on the new formulation, an algorithm to estimate the best head pose is derived. The experimental results show that the proposed extended model-based 3D LKT achieves higher accuracy and reliability than the conventional one does. Particularly, the proposed LKT is very effective in handling varying illumination, which cannot be well handled in the original LKT.     

A robust method for detecting arbitrarily tilted human faces in color images

This paper presents the design of a robust face detector that can detect arbitrarily tilted human faces in color images. This detector locates face regions by identifying mouth corners and eyes. The novel techniques included in the proposed detector are: (1) a method for compensating the colors of the input images, (2) a method for deskewing tilted faces, (3) a method for locating mouth corners, and (4) a discriminant function for positioning eyes. According to the performance evaluation on three test databases which contain 1791 faces on 1580 images, the proposed method achieves a precision rate of 94.62% and a recall rate of 92.24% in average at the detection speed of 1.6 faces per second. The performance of the proposed detector also slightly outperforms a detector from CMU.     

4-Methyl-2,4-bis(4-hydroxyphenyl)pent-1-ene,a Major Active Metabolite of Bisphenol A,Triggers Pancreatic β-Cell Death via a JNK/AMPKα Activation-Regulated Endoplasmic Reticulum Stress-Mediated Apoptotic Pathway

4-methyl-2,4-bis(4-hydroxyphenyl)pent-1-ene (MBP), a major active metabolite of bisphenol A (BPA), is generated in the mammalian liver. Some studies have suggested that MBP exerts greater toxicity than BPA. However, the mechanism underlying MBP-induced pancreatic β-cell cytotoxicity remains largely unclear. This study demonstrated the cytotoxicity of MBP in pancreatic β-cells and elucidated the cellular mechanism involved in MBP-induced β-cell death. Our results showed that MBP exposure significantly reduced cell viability, caused insulin secretion dysfunction, and induced apoptotic events including increased caspase-3 activity and the expression of active forms of caspase-3/-7/-9 and PARP protein. In addition, MBP triggered endoplasmic reticulum (ER) stress, as indicated by the upregulation of GRP 78, CHOP, and cleaved caspase-12 proteins. Pretreatment with 4-phenylbutyric acid (4-PBA; a pharmacological inhibitor of ER stress) markedly reversed MBP-induced ER stress and apoptosis-related signals. Furthermore, exposure to MBP significantly induced the protein phosphorylation of JNK and AMP-activated protein kinase (AMPK)α. Pretreatment of β-cells with pharmacological inhibitors for JNK (SP600125) and AMPK (compound C), respectively, effectively abrogated the MBP-induced apoptosis-related signals. Both JNK and AMPK inhibitors also suppressed the MBP-induced activation of JNK and AMPKα and of each other. In conclusion, these findings suggest that MBP exposure exerts cytotoxicity on β-cells via the interdependent activation of JNK and AMPKα, which regulates the downstream apoptotic signaling pathway.     

Effect of hydrothermal treatment on the performance of RuO2-Ta2O5/Ti electrodes for use in supercapacitors

Hydrothermal treatment (HTT) of RuO₂-Ta₂O₅/Ti electrode, as a method for improving their performance, for use in supercapacitors was investigated.The results show that HTT significantly enhances the stability of the electrodes. The specific capacitance of electrodes, subject to HTT in the temperature range 180-250 °C remains unchanged after 1000 CV cycles between −0.2 and 1.1 V vs. SCE; without HTT a decay to 97% of the initial is observed. The results also show that HTT decreases the activity of the electrodes for O₂ and H₂ evolution and increases the voltage window by 56-135 mV for supercapacitors, but with a specific capacitance decrease of 7-27%. XPS analyses show the existence of more hydroxides after the HTT, which leads to a little increase in the interplanar distance as indicated in the XDR results. Contact angle measurements show the presence of a more hydrophilic surface after HTT.     

Self-patterning of high-performance thin film transistors

We have developed a technique for the preparation of thin film transistors (TFTs) through the self-patterning of various organic and inorganic materials via solution processing using a wide range of solvents. To obtain selectively self-patterned layers, we treated the oxide dielectric with two-phase patterned self-assembled monolayers of hexamethyldisilazane (HMDS) and octyltrichlorosilane. The conducting polymer poly(3,4-ethylenedioxythiophene) doped with poly(styrene sulfonic acid) in water and the dielectric polymer poly(vinyl phenol) in propylene glycol methyl ether acetate were both selectively deposited and patterned on the HMDS regions with high-quality feature shapes. When source and drain electrodes were patterned on the bottom-gate oxide wafer, we also self-patterned organic and inorganic semiconductors around the channel (HMDS) regions. These TFT devices exhibited moderate to good electronic characteristics. This method has great potential for the economical full solution processing of large-area electronic devices. The selectivity in the patterning phenomena can be understood in terms of surface energy interactions.     

Stator design of a new type of spherical piezoelectric motor

The stator design of a new type of spherical motor driven by piezoelectric actuators is developed. A curved piezoelectric actuator is designed to attach to the spherical surface. A series of the curved piezoelectric actuators is laid in a line around a spherical surface. By applying an appropriate voltage signal with phase difference on neighboring actuators, a traveling wave is generated on the hemispherical shell. Each set of curved piezoelectric actuators is designed to provide motion with a single degree-of-freedom (DOF). With two or three sets of the piezoelectric actuators constructed to be mutually perpendicular, the motor can provide 2-DOF or 3-DOF motion. Stator design and analysis and experiment for the 1-, 2-, and 3-DOF conditions are presented in this article. Analytical calculation and experiment results of several fundamental characteristics of the stator are in good agreement. Performance evaluation of rotation speed and torque of the stator and some implementation problems are also addressed.     

Using multithreshold quadratic sigmoidal neurons to improveclassification capability of multilayer perceptrons

This letter proposes a new type of neurons called multithreshold quadratic sigmoidal neurons to improve the classification capability of multilayer neural networks. In cooperation with single-threshold quadratic sigmoidal neurons, the multithreshold quadratic sigmoidal neurons can be used to improve the classification capability of multilayer neural networks by a factor of four compared to committee machines and by a factor of two compared to the conventional sigmoidal multilayer perceptrons.     

Divide-and-conquer learning and modular perceptron networks

A novel modular perceptron network (MPN) and divide-and-conquer learning (DCL) schemes for the design of modular neural networks are proposed. When a training process in a multilayer perceptron falls into a local minimum or stalls in a flat region, the proposed DCL scheme is applied to divide the current training data region into two easier to be learned regions. The learning process continues when a self-growing perceptron network and its initial weight estimation are constructed for one of the newly partitioned regions. Another partitioned region will resume the training process on the original perceptron network. Data region partitioning, weight estimating and learning are iteratively repeated until all the training data are completely learned by the MPN. We evaluated and compared the proposed MPN with several representative neural networks on the two-spirals problem and real-world dataset. The MPN achieved better weight learning performance by requiring much less data presentations during the network training phases, and better generalization performance, and less processing time during the retrieving phase.     

Independent query refinement and feature re-weighting using positive and negative examples for content-based image retrieval

Query refinement and feature re-weighting are the two core techniques underlying the relevance feedback of content-based image retrieval. Most existing relevance feedback mechanisms generally model the user’s query target with a single query point and weight each extracted feature with a single importance factor. A designed estimation procedure then estimates the best query point and all importance factors by optimizing a formulated criterion which measures the goodness of the estimation. This formulated criterion simultaneously encapsulates all positive and negative examples supplied from the user’s feedback. Under such formulation, the positive and negative examples may contribute contradictorily to the criterion and sometimes may introduce higher difficulty in attaining a good estimation. In this paper, we propose a different statistical formulation to estimate independently two pairs of query points and feature weights from the positive examples and negative examples, respectively. These two pairs then define the likelihood ratio, a criterion term used to rank the relevance of all database images. This approach simplifies the criterion formulation and also avoids the mutual impeditive influence between positive examples and negative examples. The experimental results demonstrate that the proposed approach outperforms some other related approaches.

A hybrid approach of neural networks and grey modeling for adaptive electricity load forecasting

This paper proposes an effective fusion of neural networks and grey modeling for adaptive electricity load forecasting. The fusion employs the complementary strength of these two appealing techniques. In terms of forecasting accuracy, the proposed fusion scheme outperforms the individual ones and the statistical autoregressive methods according to the results of a substantial number of experiments. In addition to the fusion scheme, this paper also proposes a grey relational analysis to automatically assess the importance of each input variable for the forecasting task. This analysis helps the forecaster choose dominant ones among the many input variables, thus removing much burden of acquiring professional domain knowledge for problems and reducing the interference of irrelevant inputs on the forecasting. Experimental results are shown in this paper to verify the effectiveness of the grey relational analysis.