Intermediate-State HMMs to Capture Continuously-Changing Signal Features

机译：中间状态HMM捕获连续变化的信号特征

获取原文

页面导航

摘要
著录项
相似文献
相关主题

摘要

Traditional discrete-state HMMs are not well suited for describing steadily evolving, path-following natural processes like motion capture data or speech. HMMs cannot represent incremental progress between behaviors, and sequences sampled from the models have unnatural segment durations, unsmooth transitions, and excessive rapid variation. We propose to address these problems by permitting the state variable to occupy positions between the discrete states, and present a concrete left-right model incorporating this idea. We call this intermediate-state HMMs. The state evolution remains Markovian. We describe training using the generalized EM-algorithm and present associated update formulas. An experiment shows that the intermediate-state model is capable of gradual transitions, with more natural durations and less noise in sampled sequences compared to a conventional HMM.

机译：传统的离散状态HMM不太适合描述平稳发展的，遵循路径的自然过程，例如运动捕获数据或语音。 HMM不能表示行为之间的增量进度，并且从模型中采样的序列具有不自然的段持续时间，不平滑的过渡和过度的快速变化。我们建议通过允许状态变量占据离散状态之间的位置来解决这些问题，并提出一个包含此思想的具体左右模型。我们称此为中间状态HMM。状态演化仍然是马尔可夫式的。我们描述了使用广义EM算法的训练，并介绍了相关的更新公式。实验表明，与传统的HMM相比，中间状态模型能够逐步过渡，具有更长的自然持续时间和更少的采样序列噪声。

著录项

来源
《Annual conference of the International Speech Communication Association;INTERSPEECH 2011》|2011年|p.1828-1831|共4页
会议地点
作者
Gustav Eje Henter; W. Bastiaan Kleijn;
展开▼
作者单位

展开▼
会议组织
原文格式 PDF
正文语种
中图分类通信;
关键词
markov models; HMMs; speech synthesis;

机译：马可夫模型; HMM;语音合成;

相似文献

外文文献
中文文献
专利

1. Arabic Speech Recognition System Based on MFCC and HMMs [J] . Hussien A. Elharati, Mohamed Alshaari, Veton Z. K?puska 电脑和通信（英文） . 2020,第003期
2. Identifying anomalous signals in GPS data using HMMs: An increased likelihood of earthquakes? [J] . Wang T., Bebbington M. Computational statistics & data analysis . 2013,第Null期

机译：使用HMM识别GPS数据中的异常信号：地震的可能性增加了吗？
3. A Novel Vector Representation of Stochastic Signals Based on Adapted Ergodic HMMs [J] . Tang H., Hasegawa-Johnson M., Huang T. Signal Processing Letters, IEEE . 2010,第8期

机译：基于自适应遍历HMM的随机信号向量表示
4. Advances in classification of EEG signals via evolving fuzzy classifiers and dependant multiple HMMs. [J] . Xydeas C, Angelov P, Chiao SY, Computers in Biology and Medicine . 2006,第10期

机译：通过发展的模糊分类器和相关的多个HMM在脑电信号分类方面的进展。
5. Regularization using geometric information between sensors capturing features from brain signals [C] . Higashi, Hiroshi IEEE International Conference on Acoustics, Speech and Signal Processing;ICASSP . 2012

机译：使用传感器之间的几何信息进行正则化以捕获来自大脑信号的特征
6. Development and optimization of a HMM-ANN hybrid structure for robust recognition of impaired speech signals [D] . Polur, Prasad Babu Deenadayalan 2005

机译：HMM-ANN混合结构的开发和优化，可用于语音信号的鲁棒识别
7. HMMPred: Accurate Prediction of DNA-Binding Proteins Based on HMM Profiles and XGBoost Feature Selection [O] . Xiuzhi Sang, Wanyue Xiao, Huiwen Zheng, 2020

机译：HMMPred：基于HMM轮廓和XGBoost特征选择的DNA结合蛋白的准确预测
8. Advances in classification of EEG signals via evolving fuzzy classifiers and dependant multiple HMMs. [O] . Xydeas Costas, Angelov Plamen, Chiao Shih-Yang, 2006

机译：通过发展的模糊分类器和相关的多个HMM在脑电信号分类方面的进展。
9. Integrated Models of Signals and Background for an HMM/Neural Net Ocean AcousticEvents Classifier [R] . Huang, W. Y., Rose, R. C. 1991

机译：Hmm /神经网络海洋声学事件分类器的信号和背景的集成模型

Intermediate-State HMMs to Capture Continuously-Changing Signal Features

摘要

著录项

相似文献

相关主题

期刊订阅