Support NNEF execution model for NNAPI

The Journal of Supercomputing - With growing applications such as image recognition, speech recognition, ADAS, and AIoT, artificial intelligence (AI) frameworks are becoming popular in various...     

Secretome analysis of novel IgE-binding proteins from Penicillium citrinum

The Penicillium genus of fungi is a frequently reported cause of allergic reactions. However, only a limited number of allergens have been reported. In Penicillium spp., many allergens show higher IgE-binding activity in culture filtrate extracts than in cellular extracts. In order to investigate the IgE-reactive profile of mold-sensitized patients, secreted IgE-reactive proteins from Penicillium citrinum were identified by 2-DE, serum immunoblotting, and nanoLC-MS/MS. Among the IgE-reactive spots, one known allergen, Pen c 13, and four novel allergens were identified. The cDNAs coding for Pen c 32 and Pen c 30 were cloned using designed primers based on nanoLC-MS/MS analysis. The amino acid sequences of Pen c 32 and Pen c 30 were, respectively, found to have extensive similarity with those of pectate lyases and catalases from various fungi. Native Pen c 30 was shown to have catalase activity and to bind to serum IgE from 48% of mold-allergic patients and induced immediate type skin reactions in a sensitized patient. Here, we present a proteome approach which resulted in the identification of four novel secreted allergens. These novel allergens might be useful in allergy diagnosis and in the treatment of mold-allergic disorders.     

An output-based adaptive iterative learning controller for high relative degree uncertain linear systems

In this paper, we derive an output tracking error model based on signals filtered from plant input and output, and then present a new output-based adaptive iterative learning controller for repeatable linear systems with unknown parameters, high relative degree, initial resetting error, input disturbance and output noise. The proposed controller solves the important robustness issues without assuming the bounds of uncertainties to be sufficiently small and can be applied to high relative degree plants without using output differentiation. Control parameters are updated between successive iterations so as to compensate for unknown system parameters and uncertainties. It is shown that the internal signals inside closed-loop learning system remain bounded and the output tracking error will asymptotically converge to a profile tunable by some design parameters. Furthermore, the learning speed is easily improved if the learning gain is increased.     

Iterative learning of model reference adaptive controller for uncertain nonlinear systems with only output measurement

In this paper, a model reference adaptive control strategy is used to design an iterative learning controller for a class of repeatable nonlinear systems with uncertain parameters, high relative degree, initial output resetting error, input disturbance and output noise. The class of nonlinear systems should satisfy some differential geometric conditions such that the plant can be transformed via a state transformation into an output feedback canonical form. A suitable error model is derived based on signals filtered from plant input and output. The learning controller compensates for the unknown parameters, uncertainties and nonlinearity via projection type adaptation laws which update control parameters along the iteration domain. It is shown that the internal signals remain bounded for all iterations. The output tracking error will converge to a profile which can be tuned by design parameters and the learning speed is improved if the learning gain is large.     

A novel method for determining machine subgroups and backups with an empirical study for semiconductor manufacturing

Wafer fabrication for semiconductor manufacturing consists of multiple layers, in which the displacements (i.e., overlay errors) between layers should be reduced to enhance the yield. Although it can reduce variance between layers by fixing the exposure machine (i.e. steeper or scanner), it is not practical to expose the wafer on the same machine from layer to layer for the lengthy fabrication process in real setting. Thus, there is a critical need to determine the similarity machine subgroups, in which appreciate backups for unexpected machine down can be also prioritized. This study aims to develop a novel methodology to fill this gap based on the proposed similarity measurement of systematic overlay errors and residuals. The proposed methodology was validated via empirical study in a wafer fab and the results showed practical viability of this approach. Received: May 2005 / Accepted: December 2005     

Constraint-based Correspondence Matching for Stereo-based Interactive Robotic Massage Machine

Locating the 3D positions of the points on the human back is an essential issue in stereo-based interactive robotic back massage machines. In stereoscopic 3D localization, the 3D positions are determined from the corresponding image points captured by calibrated stereo cameras. However, detecting these corresponding points on the human back is highly challenging due to the smooth and texture-less characteristics of human skin. In the present study, this problem is resolved by means of a novel correspondences detection scheme designated as Correspondences from Epipolar geometry and Contours via Triangle barycentric coordinates (CECT). In the proposed approach, reliable correspondences are extracted from the edge contours of the human back by applying epipolar geometry, and these correspondences are then used to compute the correspondences of the featureless points within the edge contour using a triangle barycentric coordinate approach. The accuracy and robustness of the estimated correspondences are ensured by applying three geometric constraints, namely a similarity constraint, a shape constraint and an epipolar constraint. The performance of the proposed approach is demonstrated by means of a series of experiments involving 28 subjects and four different testing conditions. In addition, the accuracy of the proposed localization scheme is evaluated by comparing the estimated 3D positions with those obtained using the cun-based measurement method in Traditional Chinese Medicine (TCM).     

Inactivation of Microorganisms on the Photocatalytic Surfaces in Air

Photocatalysis by TiO 2 is a new technology that can generate strong oxidant radicals to react with organic pollutants and sterilize water. In this research, the disinfections of two microorganisms such as Bacillus subtilis ( B. subtilis ) and Penicillium citrinum ( P. citrinum ) were investigated. The commercial TiO 2 filter and immobilized TiO 2 slide were used as photocatalytic surfaces. The light intensity levels included 240, 740, 1400, and 2100 w W/cm 2 . Our results indicated that higher black light intensity resulted in higher microorganism inactivation rates. It was demonstrated that not only photocatalysis but also photolysis had germicidal effects on P. citrinum in both kinds of surface matrixes. For B. subtilis , photolysis was more significant than photocatalysis. In addition, microorganism inactivation rates of the TiO 2 filter were found to be smaller than those of the TiO 2 slide for both B. subtilis and P. citrinum . At a light intensity of 740 w W/cm 2 , the calculated 50% survival inactivation time of the TiO 2 filter for B. subtilis and P. citrinum were 0.99 h and 2.55 h, respectively. The calculated 50% survival inactivation time of the TiO 2 slide for B. subtilis and P. citrinum were 0.61 h and 1.64 h, respectively. Furthermore, the inactivation rate of P. citrinum was smaller than B. subtilis on both filter and slide. This might be related to the fact that P. citrinum was more resistant than B. subtilis .     

Digital preservation of a Taiwanese historical settlement: Using 3D post-construction scan to develop an application framework and reference for Beipu Township

The purpose of this study is to build a post-construction 3D computer model of the historical settlement of Beipu Township in Taiwan as an application and reference framework. The cultural assets were digitally preserved in a post-construction and modified form as an attempt to fulfill the life-cycle management of data for culture workers, researchers and architects. The scan process was managed in four hierarchies: the whole region, the seven historical buildings, the four streets and the special features, in different scales. The final digital model integrates research results from all involved parties, integrates local practice needs, represents hypothesis in city development, enhances historical understanding and interpretation, and facilitates remote access.     

An effective algorithm for numerical Schrödinger solver of quantum well structures

An effective algorithm for a one dimensional Schrödinger solver is proposed. The algorithm is derived from Bohm’s form of Schrödinger equation, and can be interpreted as a generalization of the Numerov process for the computational solution to the Schrödinger equation with quantum well structures. The proposed algorithm averages the probability slope’s discontinuity to converge the given energy to the nearest eigenstate, and generate complete eigenstates in a general heterojunction potential well. The new algorithm is applied to double-well III–V nanoscale MOSFET and generates each of the quantized levels.