LaViola Joseph J, Zeleznik Robert C
School of Electrical Engineering and Computer Science, University of Central Florida, Engineering 3--Harris Center, Orlando, FL 32816-2362, USA.
IEEE Trans Pattern Anal Mach Intell. 2007 Nov;29(11):1917-26. doi: 10.1109/TPAMI.2007.1109.
We present a practical technique for using a writer-independent recognition engine to improve the accuracy and speed while reducing the training requirements of a writer-dependent symbol recognizer. Our writer-dependent recognizer uses a set of binary classifiers based on the AdaBoost learning algorithm, one for each possible pairwise symbol comparison. Each classifier consists of a set of weak learners, one of which is based on a writer-independent handwriting recognizer. During online recognition, we also use the n-best list of the writer-independent recognizer to prune the set of possible symbols and thus reduce the number of required binary classifications. In this paper, we describe the geometric and statistical features used in our recognizer and our all-pairs classification algorithm. We also present the results of experiments that quantify the effect incorporating a writer-independent recognition engine into a writer-dependent recognizer has on accuracy, speed, and user training time.
我们提出了一种实用技术，即使用独立于书写者的识别引擎来提高准确性和速度，同时降低依赖于书写者的符号识别器的训练要求。我们的依赖于书写者的识别器使用基于AdaBoost学习算法的一组二元分类器，每个可能的成对符号比较对应一个分类器。每个分类器由一组弱学习器组成，其中一个基于独立于书写者的手写识别器。在在线识别过程中，我们还使用独立于书写者的识别器的n-best列表来修剪可能的符号集，从而减少所需的二元分类数量。在本文中，我们描述了我们的识别器中使用的几何和统计特征以及我们的全对分类算法。我们还展示了实验结果，这些实验量化了将独立于书写者的识别引擎纳入依赖于书写者的识别器对准确性、速度和用户训练时间的影响。
