Adaptive Learning of Classifiers and Its Applications

分类器的自适应学习及其应用

• 批准号: 15500088
• 负责人: MARUYAMA Minoru
• 金额: $ 1.6万
• 依托单位: Shinshu University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (C)
• 财政年份: 2003
• 资助国家: 日本
• 起止时间: 2003 至 2004
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-15500088/
• 关键词: learning from examples; SVM; boosting; radial basis function; approximation; classification; virtual examples; 学習; サンプリング; 画像理解; 時系列データ; RBF; スパース化; 画像認識

To recognize target objects in the image, technique of learning from examples are often exploited. To learn the good classifier, usually a lot of samples are required. On the other hand, human beings can recognize an object efficiently even if it is a novel object. To realize such ability on machines, it is necessary to develop a learning method to give classifiers which are efficient with respect to both recognition rate and recognition speed, from very small learning examples. Classifiers based on the non-linear functions are often used to recognize pattern in a image. Although their classification performance is quite good, their computational cost is often prohibitive. To overcome the difficulty, in this report, we first propose a method to recognize patterns in a image based on the hierarchical classifiers. It consists of two kinds of classifiers : simple and fast classifier, whose performance is not necessarily satisfactory, and the non-linear slow classifier, whose performance is quite good. With the use of first (bottom) classifier, many false targets are rejected. The second, complex classifier is applied to the remaining patterns. By this architecture, fast and reliable classifiers are made possible. As the non-linear classifiers, we use RBF-SVM. In the case of SVM, recognition speed depends on the number of support vectors. To improve the recognition speed further, we propose a method to approximate the SVM based on the sampling from the support vectors. We also propose methods to learn fast and reliable classifiers based on the boosting algorithm subject to the upper bound of the number of RBF centers. To learn the classifiers from the limited number of examples, we propose a method to utilize virtual examples. In the report, we first propose the synthesis method based on the simple geometric transformation. Then, to improve the quality of the samples for hand written pattems, we also propose amethod to generate pattems from online data.

为了识别图像中的目标对象，经常利用从示例中学习的技术。为了学习好的分类器，通常需要大量的样本。另一方面，即使是新物体，人类也可以有效地识别该物体。为了在机器上实现这种能力，有必要开发一种学习方法，从非常小的学习示例中给出在识别率和识别速度方面都有效的分类器。基于非线性函数的分类器通常用于识别图像中的模式。尽管它们的分类性能相当好，但它们的计算成本往往令人望而却步。为了克服这个困难，在本报告中，我们首先提出了一种基于分层分类器来识别图像中的模式的方法。它由两种分类器组成：简单快速分类器，其性能不一定令人满意，以及非线性慢速分类器，其性能相当不错。通过使用第一个（底部）分类器，许多错误目标被拒绝。第二个复杂的分类器应用于其余模式。通过这种架构，快速可靠的分类器成为可能。作为非线性分类器，我们使用 RBF-SVM。对于SVM，识别速度取决于支持向量的数量。为了进一步提高识别速度，我们提出了一种基于支持向量采样来近似 SVM 的方法。我们还提出了基于 RBF 中心数量上限的 boosting 算法来学习快速可靠的分类器的方法。为了从有限数量的示例中学习分类器，我们提出了一种利用虚拟示例的方法。在报告中，我们首先提出了基于简单几何变换的合成方法。然后，为了提高手写模式样本的质量，我们还提出了一种从在线数据生成模式的方法。

项目成果

Off-line handwritten character recognition by SVM based on the virtual examples synthesized from on-line characters

基于在线字符合成虚拟样例的SVM离线手写字符识别

• 期刊: Proc.International Conference on Document Analysis and Recognition 2005 (ICDAR2005) (to appear)
• 作者: M.Maruyama;H.Miyao;Y.Nakano
• 通讯作者: Y.Nakano

An online handwritten music score recognition system

在线手写乐谱识别系统

• 发表时间: 2004
• 期刊: Proc.Int'l Conf.on Pattern Recognition 2004
• 作者: H.Miyao;M.Maruyama
• 通讯作者: M.Maruyama

階層型識別器を用いた情景画像からの文字抽出手法

使用层次分类器的场景图像字符提取方法

• 发表时间: 2005
• 期刊: 電子情報通信学会論文誌DII (印刷中)
• 作者: 山口拓磨;丸山稔
• 通讯作者: 丸山稔

M.Maruyama, T.Yamaguchi: "Character extraction from natural scene images by hierarchical classifiers"Proc.Int'l Conference on Pattern Recognition 2004. (accepted). (2004)

M.Maruyama、T.Yamaguchi：“通过层次分类器从自然场景图像中提取字符”Proc.Intl Conference on Pattern Recognition 2004。（已接受）。

Character extraction from natural scene images by hierarchical classifiers

• DOI: 10.1109/icpr.2004.1334352
• 发表时间: 2004-08
• 期刊: Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004.
• 作者: Takuma Yamaguchi;M. Maruyama