[论文解读] Large Vocabulary Arabic Online Handwriting Recognition System

Ibrahim Abdelaziz, Sherif Abdou+1 more|arXiv (Cornell University)|Oct 17, 2014

Handwritten Text Recognition Techniques参考文献 35被引用 9

一句话总结

本论文提出了一种基于HMM的大词汇量在线阿拉伯文手写识别系统，支持64,000个词汇（覆盖92%的语言），采用先进的语音识别启发式建模技术及一种新颖的预处理方法以处理延迟笔画。该系统在每位用户使用少于200个词进行写作者自适应的情况下，实现了87.5%的准确率，无论在小词汇量还是大词汇量设置下，其性能均优于当前最先进方法，并具备实时处理能力。

ABSTRACT

Arabic handwriting is a consonantal and cursive writing. The analysis of Arabic script is further complicated due to obligatory dots/strokes that are placed above or below most letters and usually written delayed in order. Due to ambiguities and diversities of writing styles, recognition systems are generally based on a set of possible words called lexicon. When the lexicon is small, recognition accuracy is more important as the recognition time is minimal. On the other hand, recognition speed as well as the accuracy are both critical when handling large lexicons. Arabic is rich in morphology and syntax which makes its lexicon large. Therefore, a practical online handwriting recognition system should be able to handle a large lexicon with reasonable performance in terms of both accuracy and time. In this paper, we introduce a fully-fledged Hidden Markov Model (HMM) based system for Arabic online handwriting recognition that provides solutions for most of the difficulties inherent in recognizing the Arabic script. A new preprocessing technique for handling the delayed strokes is introduced. We use advanced modeling techniques for building our recognition system from the training data to provide more detailed representation for the differences between the writing units, minimize the variances between writers in the training data and have a better representation for the features space. System results are enhanced using an additional post-processing step with a higher order language model and cross-word HMM models. The system performance is evaluated using two different databases covering small and large lexicons. Our system outperforms the state-of-art systems for the small lexicon database. Furthermore, it shows promising results (accuracy and time) when supporting large lexicon with the possibility for adapting the models for specific writers to get even better results.

研究动机与目标

• 为解决在大型词典下在线手写系统中识别连笔且形态丰富的阿拉伯文文本的挑战。
• 克服现有系统在处理延迟笔画及小词汇量支持方面存在的局限性。
• 开发一种可扩展、实时的识别系统，能够在使用极少用户特定适应数据的情况下实现高准确率。
• 将语音识别中先进的HMM训练技术整合到阿拉伯文手写识别中，以提升性能。

提出的方法

• 系统采用具有上下文相关状态的隐马尔可夫模型（HMM）及高斯混合模型，实现鲁棒的序列建模。
• 提出一种新颖的预处理技术，在无需初始检测延迟笔画的情况下重新排序笔画，从而更好地与词结构对齐。
• 采用的高级训练技术包括说话人自适应训练、判别式微调以及自适应高斯混合模型拆分。
• 对实时笔画序列（x, y, 时间）进行特征提取，生成HMM的输入。
• 识别采用两阶段解码策略：第一阶段使用64,000词汇词典，第二阶段使用五元语法语言模型对假设进行重打分。
• 通过每位用户少于200个词的写作者特定适应数据，实现模型微调，显著提升准确率。

实验结果

研究问题

• RQ1 如何设计一种支持64,000个唯一词汇且具备高准确率的大词汇量阿拉伯文在线手写识别系统？
• RQ2 何种预处理方法能有效处理不同书写风格下的延迟笔画，且无需检测笔画？
• RQ3 语音识别中先进的HMM训练技术在多大程度上可提升阿拉伯文手写识别的性能？
• RQ4 仅使用极少用户特定适应数据（少于200词）是否能显著提升个体写作者的识别准确率？
• RQ5 在大词汇量环境下，高阶语言模型的集成对识别准确率有何影响？

主要发现

• 在首次解码后，使用写作者无关模型在ALTEC-AH测试集上达到80.07%的准确率，经第二轮使用五元语法语言模型重打分后，准确率提升至87.47%。
• 仅使用每位写作者200个词的适应数据，系统即达到87.47%的准确率，展现出强大的个性化能力。
• 对于词典内的词汇，适应后系统准确率达到95%，表明对已知词汇库条目的高度可靠性。
• 系统保持了接近实时的性能，小词汇量样本识别时间低于1秒，大词汇量样本识别时间约为2.2秒。
• 仅耗时0.15秒的第二轮重打分显著提升了识别结果，证明了高阶语言建模的有效性。
• 该系统在小词汇量基准测试中优于现有最先进方法，并在大词汇量应用中展现出巨大潜力。