On the use of small training sets for neural network-based characterization of mixed pixels in remotely sensed hyperspectral images

In this work, neural network-based models involved in hyperspectral image spectra separation are considered. Focus is on how to select the most highly informative samples for effectively training the neural architecture. This issue is addressed here by several new algorithms for intelligent selection of training samples: (1) a border-training algorithm (BTA) which selects training samples located in the vicinity of the hyperplanes that can optimally separate the classes; (2) a mixed-signature algorithm (MSA) which selects the most spectrally mixed pixels in the hyperspectral data as training samples; and (3) a morphological-erosion algorithm (MEA) which incorporates spatial information (via mathematical morphology concepts) to select spectrally mixed training samples located in spatially homogeneous regions. These algorithms, along with other standard techniques based on orthogonal projections and a simple Maximin-distance algorithm, are used to train a multi-layer perceptron (MLP), selected in this work as a representative neural architecture for spectral mixture analysis. Experimental results are provided using both a database of nonlinear mixed spectra with absolute ground truth and a set of real hyperspectral images, collected at different altitudes by the digital airborne imaging spectrometer (DAIS 7915) and reflective optics system imaging spectrometer (ROSIS) operating simultaneously at multiple spatial resolutions.     

Parameter estimation of Poisson mixture with automated model selection through BYY harmony learning

Finite mixture is widely used in the fields of information processing and data analysis. However, its model selection, i.e., the selection of components in the mixture for a given sample data set, has been still a rather difficult task. Recently, the Bayesian Ying-Yang (BYY) harmony learning has provided a new approach to the Gaussian mixture modeling with a favorite feature that model selection can be made automatically during parameter learning. In this paper, based on the same BYY harmony learning framework for finite mixture, we propose an adaptive gradient BYY learning algorithm for Poisson mixture with automated model selection. It is demonstrated well by the simulation experiments that this adaptive gradient BYY learning algorithm can automatically determine the number of actual Poisson components for a sample data set, with a good estimation of the parameters in the original or true mixture where the components are separated in a certain degree. Moreover, the adaptive gradient BYY learning algorithm is successfully applied to texture classification.     

A compiler for parsing schemata

We present a compiler that can be used to automatically obtain efficient Java implementations of parsing algorithms from formal specifications expressed as parsing schemata. The system performs an analysis of the inference rules in the input schemata in order to determine the best data structures and indexes to use, and to ensure that the generated implementations are efficient. The system described is general enough to be able to handle all kinds of schemata for different grammar formalisms, such as context‐free grammars and tree‐adjoining grammars, and it provides an extensibility mechanism allowing the user to define custom notational elements. This compiler has proven very useful for analyzing, prototyping and comparing natural‐language parsers in real domains, as can be seen in the empirical examples provided at the end of the paper. Copyright © 2008 John Wiley & Sons, Ltd.     

Testing a MODIS Global Disturbance Index across North America

Large-scale ecosystem disturbances (LSEDs) have major impacts on the global carbon cycle as large pulses of CO₂ and other trace gases from terrestrial biomass loss are emitted to the atmosphere during disturbance events. The high temporal and spatial variability of the atmospheric emissions combined with the lack of a proven methodology to monitor LSEDs at the global scale make the timing, location and extent of vegetation disturbance a significant uncertainty in understanding the global carbon cycle. The MODIS Global Disturbance Index (MGDI) algorithm is designed for large-scale, regular, disturbance mapping using Aqua/Moderate Resolution Imaging Spectroradiometer (MODIS) Land Surface Temperature (LST) and Aqua/MODIS Enhanced Vegetation Index (EVI) data. The MGDI uses annual maximum composite LST data to detect fundamental changes in land-surface energy partitioning, while avoiding the high natural variability associated with tracking LST at daily, weekly, or seasonal time frames. Here we apply the full Aqua/MODIS dataset through 2006 to the improved MGDI algorithm across the woody ecosystems of North America and test the algorithm by comparison with confirmed, historical wildfire events and the windfall areas of documented major hurricanes. The MGDI accurately detects the location and extent of wildfire throughout North America and detects high and moderate severity impacts in the windfall area of major hurricanes. We also find detections associated with clear-cut logging and land-clearing on the forest-agricultural interface. The MGDI indicates that 1.5% (195,580 km²) of the woody ecosystems within North America was disturbed in 2005 and 0.5% (67,451 km²) was disturbed in 2006. The interannual variability is supported by wildfire detections and official burned area statistics.     

METAPHOR RECOGNITION: CHMETA, A PATTERN-BASED SYSTEM

Metaphor recognition presents a computational challenge, in part due to metaphoric deviation from literal thinking, and also because of a metaphor's various linguistic expressions. This article forwards a new computational method, an integrated treatment of metaphor recognition from the computational perspective, which recent related studies have not entirely addressed. The authors differentiate metaphor recognition from complex metaphor inference and interpretation employing psychological clues. To accomplish this, we have developed a formalized system of metaphorical expression in metaphor role dependency schema, which specifically defines, classifies, and quantifies metaphorical anomalies, building a computable classification system for metaphors (incorporating 32 major patterns of metaphorical expressions) by providing a strategy to locate potential metaphorical anomalies in a target input sentence through a pattern recognition method and a metaphor components' tagging approach. This metaphor recognition and tagging system is named and implemented as "CHMeta." Experiment results support the validity and efficiency of this metaphor recognition system. Compared with most metaphor computation systems, which work mainly on a few examples, this system classifies major metaphorical expressions from a computational perspective and is able to recognize a variety of different kinds of metaphors, including nested ones. Thus, this is the first integrated work in computable classification, recognition, and tagging of large-scale metaphors in Chinese.     

Global asymptotic stability of delayed neural networks with discontinuous neuron activations

This paper investigates a class of delayed neural networks whose neuron activations are modeled by discontinuous functions. By utilizing the Leray–Schauder fixed point theorem of multivalued version, the properties of M-matrix and generalized Lyapunov approach, we present some sufficient conditions to ensure the existence and global asymptotic stability of the state equilibrium point. Furthermore, the global convergence of the output solutions are also discussed. The assumptive conditions imposed on activation functions are allowed to be unbounded and nonmonotonic, which are less restrictive than previews works on the discontinuous or continuous neural networks. Hence, we improve and extend some existing results of other researchers. Finally, one numerical example is given to illustrate the effectiveness of the criteria proposed in this paper.     

Fault-tree analysis for liquefied natural gas terminal emergency shutdown system

Natural gas, one of the cleanest, most efficient and useful of all energy sources, is a vital component of the world’s supply of energy. To make natural gas more convenient for storage and transportation, it is refined and condensed into a liquid called liquefied natural gas (LNG). In a LNG site, safety is a long-team and critical issue. The emergency shutdown (ESD) system in the LNG receiving terminal is used to automatically stop the pumps and isolate the leakage section. Fault-tree analysis (FTA) has been widely used for providing logical functional relationships among subsystems and components of a system and identifying the root causes of the undesired failures in a system. In the conventional FTA for the ESD system, we usually assume that exact failure probabilities of events are collected. However, in most real applications, first, the FTA for the ESD system needs to be made at a early design or manufacturing stage, certain new components normally used without failure data; secondly, sometimes the environmental change in the system during the operation periods. This makes more difficult to gather past exact failures data for the FTA. To complete the FTA of the ESD system under these uncertain situations, we apply the intuitionistic fuzzy sets (IFS) theory to the FTA. We generate the intuitionistic fuzzy fault-tree interval, and the intuitionistic fuzzy reliability interval for the ESD system. We also present an algorithm to find the critical components in the system based on IFS–FTA and determine weak paths in the ESD system, where the key improvement must be made.     

一类Lyapunov型矩阵方程组的中心对称解及其最佳逼近

建立了求矩阵方程组AiXBi＋GiXDi=Fi（i=1,2）的中心对称解的迭代算法.使用该方法不仅可以判断矩阵方程组是否有中心对称解,而且在有中心对称解时,还能够在有限步迭代计算之后得到矩阵方程组的极小范数中心对称解.同时,也能够在矩阵方程组的中心对称解集合中求得给定矩阵的最佳逼近.     

基于色度坐标高斯混合模型的步态检测

针对传统的基于RGB通道的高斯混合模型低对比度像素点检测效果较差的问题,提出一种基于色度坐标的高斯混合模型,使之更好地用于步态检测。该算法将RGB色彩值转换到色度坐标上,以强调色彩对比度,提高低对比度像素点的检测率,并增加亮度信息以减小阴影的影响,在前景提取部分,加入噪声抑制机制。实验结果表明,改进后的算法在相同对比度下,误检测率最多可减小一半。     

基于多种视频特征的镜头边界检测算法

针对固定场景的视频,提出一种基于时空分割的视频分割算法.该算法在时域中利用Tophat形态学滤波得到视频对象的精确位置,在空域中采用基于t混合模型和贪婪EM的聚类算法进行单帧图像分割.将时域定位和空域分割结果结合,可以准确地将视频序列中感兴趣的运动目标分割出来.实验表明,该算法能够得到完整的视频对象,有一定的理论意义和实用性.