Exploring the use of conditional random field models and HMMs for historical handwritten document recognition
Exploring the use of conditional random field models and HMMs for historical handwritten document recognition
复制标题
DOI：
10.1109/dial.2006.19
复制
发表时间：
2006-04
期刊：
影响因子：
--
通讯作者：
Shaolei Feng;R. Manmatha;A. McCallum
中科院分区：
文献类型：
--
作者：
Shaolei Feng;R. Manmatha;A. McCallum
在本文中，我们探索了提高手写识别离散特征依赖模型性能的不同方法。隐马尔可夫模型经常用于手写识别。条件随机场（CRF）允许更一般的依赖关系，我们研究了它们的使用。我们相信这是应用CRF进行手写识别的首次尝试。我们表明，在整个单词识别任务中，CRF 在 20 页乔治·华盛顿手稿的公开标准数据集上表现优于 HMM。整个单词识别任务的规模空间很大——几乎有 1200 个状态。为了使 CRF 计算易于处理，我们使用波束搜索，通过三种不同的方法使推理更加有效。通过使用收集频率直接平滑离散特征，可以使用 HMM 获得更好的改进。这表明了平滑的重要性，也表明了当涉及大状态空间时训练 CRF 的难度
In this paper we explore different approaches for improving the performance of dependency models on discrete features for handwriting recognition. Hidden Markov models have often been used for handwriting recognition. Conditional random fields (CRF's) allow for more general dependencies and we investigate their use. We believe that this is the first attempt at apply CRF's for handwriting recognition. We show that on the whole word recognition task, the CRF performs better than a HMM on a publicly available standard dataset of 20 pages of George Washington's manuscripts. The scale space for the whole word recognition task is large - almost 1200 states. To make CRF computation tractable we use beam search to make inference more efficient using three different approaches. Better improvement can be obtained using the HMM by directly smoothing the discrete features using the collection frequencies. This shows the importance of smoothing and also indicates the difficulty of training CRF's when large state spaces are involved