The State of the Art in Text Filtering

This paper develops a conceptual framework for text filtering practice and research, and reviews present practice in the field. Text filtering is an information seeking process in which documents are selected from a dynamic text stream to satisfy a relatively stable and specific information need. A model of the information seeking process is introduced and specialized to define text filtering. The historical development of text filtering is then reviewed and case studies of recent work are used to highlight important design characteristics of modern text filtering systems. User modeling techniques drawn from information retrieval, recommender systems, machine learning and other fields are described. The paper concludes with observations on the present state of the art and implications for future research on text filtering. This revised version was published online in July 2006 with corrections to the Cover Date.     

Martha J. Farah.

Recognizes Martha J. Farah for rigorous empirical and theoretical analysis of visual cognition, in which understanding of normal function and analysis of neurological deficits illuminate and strengthen one another. Applying diverse experimental techniques within an integrated approach to the study of mind and brain, she has sharpened and helped to answer fundamental questions about how humans generate and manipulate visual images, recognize objects, and attend to positions in space. Along with a citation, a biography is presented for Farah, as well as a bibliography of her works. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

实用语音传感器

介绍了利用MCS-96单片计算机构成实用语音传感器的电路设计方案和配合该系统研究的语音识别方法以及实验结果等。     

Word recognition: From letters to meaning.

Reviews the book, Basic Processes in Reading Visual Word Recognition by Derek Besner and Glyn W. Humphreys (see record 1990-99017-000). While there has been an increase in the amount of work on many different aspects of reading, as Besner and Humphreys point out in the overview to their book, the vast majority of the research on the topic of reading in the past twenty years has been concerned with the processes involved in word recognition. For this reason, Besner and Humphreys have attempted to bring together studies on topics which are both relevant to current debates in the field of word recognition, and which are likely to be important for future developments in the field. They have compiled an edited volume consisting of their overview and eight additional chapters. The editors have attempted to span the continuum of processes involved in word recognition and thus have included chapters which cover topics ranging from the visual analysis of words, to those on the influence of semantic factors on word recognition. The authors of these chapters comprise an impressive list of researchers in the field of word recognition, with the majority of chapters being authored by leading researchers on the topic. Given the stature of the authors and the range of topics covered, in theory this volume should provide a very thorough overview of current theory and research on reading. There is no question that each of the chapters is interesting and important in its own right. However, in practice the volume as a whole fell somewhat short of my expectations. The different tacts taken by different authors has resulted in a very uneven coverage of the current debates in the field. Notwithstanding these criticisms, I am sure that the majority of researchers in this field will consider this volume to be an important contribution. The book would provide a very useful addition to graduate courses in cognitive sciences and as a supplemental text for an undergraduate course on the psychology of reading. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Learning-based dynamic coupling of discrete and continuous trackers

We present a data-driven dynamic coupling between discrete and continuous methods for tracking objects of high dofs, which overcomes the limitations of previous techniques. In our approach, two trackers work in parallel, and the coupling between them is based on the tracking error. We use a model-based continuous method to achieve accurate results and, in cases of failure, we re-initialize the model using our discrete tracker. This method maintains the accuracy of a more tightly coupled system, while increasing its efficiency. At any given frame, our discrete tracker uses the current and several previous frames to search into a database for the best matching solution. For improved robustness, object configuration sequences, rather than single configurations, are stored in the database. We apply our framework to the problem of 3D hand tracking from image sequences and the discrimination between fingerspelling and continuous signs in American Sign Language.     

一种基于稳健匹配准则的视频文本跟踪方法

本文首先根据文本中各个字符的颜色相似的特性,确定已知文本块的文本的颜色,利用文本的颜色信息二值化文本图像以得到文本块中的字符数、字符的尺寸和间距;其次,根据视频图像的大小和视频的播放速度确定搜索窗口的大小,并利用文本的运动信息来预测搜索窗口的位置;第三,根据在相邻两帧中具有相同内容文本的颜色相似的特性,提出了一个可以解决脉冲噪声和部分缺损或遮挡问题的稳健匹配准则来跟踪文本块;最后,确定跟踪输出的文本的颜色、尺寸和运动方式。本文方法可以很好地跟踪平移、缩小或放大、旋转、淡入淡出和部分被遮挡的水平或倾斜排列的文本,并且可以记录文本平移、旋转和缩放的速度。     

大庆油田杏1-3区三次加密试验区剩余油分布研究

通过对试验区近年完钻的检查井各单砂层水淹状况及其影响因素的分析。利用神经网络模式识别方法建立起单层剩余油识别模型，在此基础上对三次加密试验区各小层剩余油分布进行研究，总结出各类油层剩余油分布特征及成因类型，并分析了试验区剩余油潜力分布，从而为三次加密布井及射孔方案编制提供了可靠的依据。     

Two-tier architecture for unconstrained handwritten character recognition

In this paper, we propose an approach that combines the unsupervised and supervised learning techniques for unconstrained handwritten numeral recognition. This approach uses the Kohonen self-organizing neural network for data classification in the first stage and the learning vector quantization (LVQ) model in the second stage to improve classification accuracy. The combined architecture performs better than the Kohonen self-organizing map alone. In the proposed approach, the collection of centroids at different phases of training plays a vital role in the performance of the recognition system. Four experiments have been conducted and experimental results show that the collection of centroids in the middle of the training gives high performance in terms of speed and accuracy. The systems developed also resolve the confusion between handwritten numerals.     

检测骨架图形特征点的新方法

端点、折点、歧点和交点是骨架图形的重要特点是，提出用快速查表法检测侯选特征点，再进行侯选特征点聚集，用局部结构分析，去除假特征点，判别真特征点的类型，并对折点给以大致的定量描述。实验结果表明，该方法能快速有效地找出骨架图形的特征点，在目标识别、字符识别等应用中起着关键作用。