公开(公告)号：US20070256189A1

公开(公告)日：2007-11-01

申请号：US11380289

申请日：2006-04-26

Applicant: Jilei Tian ,  Jani Nurminen ,  Victor Popa

Inventor： Jilei Tian ,  Jani Nurminen ,  Victor Popa

IPC: C12N15/82 ,  C12N15/87

CPC classification number: G10L13/033 ,  G10L2021/0135

Abstract: Systems and methods are provided for performing soft alignment in Gaussian mixture model (GMM) based and other vector transformations. Soft alignment may assign alignment probabilities to source and target feature vector pairs. The vector pairs and associated probabilities may then be used calculate a conversion function, for example, by computing GMM training parameters from the joint vectors and alignment probabilities to create a voice conversion function for converting speech sounds from a source speaker to a target speaker.

Abstract translation: 提供了系统和方法，用于在基于高斯混合模型（GMM）和其他矢量变换中执行软对齐。 软对齐可以将对齐概率分配给源和目标特征向量对。 然后可以使用矢量对和相关联的概率来计算转换函数，例如通过从联合向量计算GMM训练参数和对齐概率来创建用于将语音从源扬声器转换为目标扬声器的语音转换功能。

公开(公告)号：US20070233625A1

公开(公告)日：2007-10-04

申请号：US11396831

申请日：2006-04-03

Applicant: Jilei Tian ,  Jani Nurminen ,  Victor Popa

Inventor： Jilei Tian ,  Jani Nurminen ,  Victor Popa

IPC: G06F15/18

CPC classification number: G06K9/628 ,  G06N99/005 ,  G10L21/007

Abstract: An apparatus for providing data clustering and mode selection includes a training element and a transformation element. The training element is configured to receive a first training data set, a second training data set and auxiliary data extracted from the same material as the first training data set. The training element is also configured to train a classifier to group the first training data set into M clusters based on the auxiliary data and the first training data set and train M processing schemes corresponding to the M clusters for transforming the first training data set into the second training data set. The transformation element is in communication with the training element and is configured to cluster the second training data set into M clusters based on features associated with the second training data set.

Abstract translation: 一种用于提供数据聚类和模式选择的装置包括训练元素和变换元素。 训练元件被配置为接收从与第一训练数据集相同的材料提取的第一训练数据集，第二训练数据集和辅助数据。 训练元素还被配置为训练分类器，以基于辅助数据和第一训练数据集以及对应于M个簇的训练M个处理方案将第一训练数据集合分组成M个群集，以将第一训练数据集变换为 第二训练数据集。 转换元件与训练元素通信，并且被配置为基于与第二训练数据集相关联的特征将第二训练数据集聚集成M个群集。

公开(公告)号：US08224648B2

公开(公告)日：2012-07-17

申请号：US11966255

申请日：2007-12-28

Applicant: Jilei Tian ,  Victor Popa ,  Jani Kristian Nurminen

Inventor： Jilei Tian ,  Victor Popa ,  Jani Kristian Nurminen

IPC: G01L13/06

CPC classification number: G10L21/00 ,  G10L2021/0135

Abstract: A hybrid approach is described for combining frequency warping and Gaussian Mixture Modeling (GMM) to achieve better speaker identity and speech quality. To train the voice conversion GMM model, line spectral frequency and other features are extracted from a set of source sounds to generate a source feature vector and from a set of target sounds to generate a target feature vector. The GMM model is estimated based on the aligned source feature vector and the target feature vector. A mixture specific warping function is generated each set of mixture mean pairs of the GMM model, and a warping function is generated based on a weighting of each of the mixture specific warping functions. The warping function can be used to convert sounds received from a source speaker to approximate speech of a target speaker.

Abstract translation: 描述了混合方法，用于组合频率扭曲和高斯混合建模（GMM），以实现更好的扬声器身份和语音质量。 为了训练语音转换GMM模型，从一组源声音中提取线谱频率和其他特征以产生源特征向量和从一组目标声音生成目标特征向量。 基于对齐的源特征向量和目标特征向量来估计GMM模型。 每个GMM模型的混合均值对都产生混合特定的翘曲函数，并且基于每个混合特定翘曲函数的加权产生翘曲函数。 翘曲功能可用于将从源扬声器接收的声音转换为目标扬声器的近似语音。

公开(公告)号：US20080147385A1

公开(公告)日：2008-06-19

申请号：US11611798

申请日：2006-12-15

Applicant: Jani Nurminen ,  Jilei Tian ,  Victor Popa

Inventor： Jani Nurminen ,  Jilei Tian ,  Victor Popa

IPC: G10L19/12

CPC classification number: G10L21/00 ,  G10L2021/0135

Abstract: An improved system method for enabling and implementing codebook-based voice conversion that both significantly reduces the memory footprint and improves the continuity of the output. In various embodiments, the paired source-target codebook is implemented as a multi-stage vector quantizer. During the conversion, N best candidates in a tree search are taken as the output from the quantizer. The N candidates for each vector to be converted are used in a dynamic programming-based approach that finds a smooth but accurate output sequence.

Abstract translation: 一种改进的系统方法，用于启用和实施基于代码本的语音转换，可显着减少内存占用并提高输出的连续性。 在各种实施例中，成对的源目标码本被实现为多级矢量量化器。 在转换期间，树搜索中的N个最佳候选者作为量化器的输出。 将要转换的每个向量的N个候选者用于基于动态规划的方法，其寻找平滑但准确的输出序列。

公开(公告)号：US20070239634A1

公开(公告)日：2007-10-11

申请号：US11400629

申请日：2006-04-07

Applicant: Jilei Tian ,  Jani Nurminen ,  Victor Popa

Inventor： Jilei Tian ,  Jani Nurminen ,  Victor Popa

IPC: G06N3/02

CPC classification number: G10L21/00 ,  G10L13/033 ,  G10L2021/0135

Abstract: An apparatus for providing efficient evaluation of feature transformation includes a training module and a transformation module. The training module is configured to train a Gaussian mixture model (GMM) using training source data and training target data. The transformation module is in communication with the training module. The transformation module is configured to produce a conversion function in response to the training of the GMM. The training module is further configured to determine a quality of the conversion function prior to use of the conversion function by calculating a trace measurement of the GMM.

Abstract translation: 用于提供特征变换的有效评估的装置包括训练模块和变换模块。 训练模块被配置为使用训练源数据和训练目标数据训练高斯混合模型（GMM）。 变换模块与训练模块通信。 转换模块被配置为响应于GMM的训练而产生转换功能。 训练模块还被配置为通过计算GMM的跟踪测量来确定在使用转换功能之前的转换功能的质量。

公开(公告)号：US20210289854A1

公开(公告)日：2021-09-23

申请号：US16823242

申请日：2020-03-18

Applicant: Ioana Livia Popa-Simil ,  Victor Popa-Simil ,  Andrei Popa-Simil

Inventor： Ioana Livia Popa-Simil ,  Victor Popa-Simil ,  Andrei Popa-Simil

IPC: A41D13/12 ,  A41D13/11 ,  A41D31/30 ,  A62B9/00 ,  A62B18/08 ,  A62B18/02

Abstract: During national crisis as virus pandemics, bio-terrorism, chemical pollution, a response is to lock down all the economy, with several % GDP loss, or to convince population to wear, reasonably, appropriate individual anti-virus and bacterial protection accessories, made of a full head and shoulders protection mask, equipped with air ventilation and processing unit, and gloves, and body protection foil, for single or multiple use, and get minimal economic losses. The air processing unit may be developed a modular structure, starting from a simple filtered fan, with air suction on top of the head, flowing air over the face, for low probability of contamination environments, upgraded to a more complex unit including heat pipe, catalytic organic matter reduction, oxygen generator and closed circuit breathing unit, that may be integrated into a full body personal protection equipment by extending the functions of temperature, humidity and pressure control to full body, for hazardous environments.

公开(公告)号：US20090094031A1

公开(公告)日：2009-04-09

申请号：US11867196

申请日：2007-10-04

Applicant: Jilei Tian ,  Victor Popa ,  Jani K. Nurminen

Inventor： Jilei Tian ,  Victor Popa ,  Jani K. Nurminen

IPC: G10L15/04

CPC classification number: G10L15/063 ,  G10L15/26 ,  G10L21/00 ,  G10L2021/0135

Abstract: An apparatus for providing text independent voice conversion may include a first voice conversion model and a second voice conversion model. The first voice conversion model may be trained with respect to conversion of training source speech to synthetic speech corresponding to the training source speech. The second voice conversion model may be trained with respect to conversion to training target speech from synthetic speech corresponding to the training target speech. An output of the first voice conversion model may be communicated to the second voice conversion model to process source speech input into the first voice conversion model into target speech corresponding to the source speech as the output of the second voice conversion model.

Abstract translation: 用于提供文本独立语音转换的装置可以包括第一语音转换模型和第二语音转换模型。 关于将训练源语音转换成对应于训练源语音的合成语音，可以对第一语音转换模型进行训练。 关于从对应于训练目标语音的合成语音到训练目标语音的转换，可以对第二语音转换模型进行训练。 第一语音转换模型的输出可以被传送到第二语音转换模型，以将输入到第一语音转换模型的源语音转换为与源语音相对应的目标语音作为第二语音转换模型的输出。

公开(公告)号：US20080082320A1

公开(公告)日：2008-04-03

申请号：US11537428

申请日：2006-09-29

Applicant: Victor Popa ,  Jani K. Nurminen ,  Jilei Tian

Inventor： Victor Popa ,  Jani K. Nurminen ,  Jilei Tian

IPC: G10L19/00

CPC classification number: G10L13/033 ,  G10L2021/0135

Abstract: An apparatus is provided that includes a converter for training a voice conversion model for converting source encoding parameters characterizing a source speech signal associated with a source voice into corresponding target encoding parameters characterizing a target speech signal associated with a target voice. To reduce the affect of noise on the voice conversion model, the converter may be configured for receiving sequences of source and target encoding parameters, and train the model without one or more frames of the source and target speech signals that have energies less than a threshold energy. After conversion of the respective parameters, then, the converter, a decoder or another component may be configured for reducing the energy of one or more frames of the target speech signal that have an energy less than the threshold energy, where the threshold value may be adaptable based upon models of speech frames and non-speech frames.

Abstract translation: 提供一种装置，其包括用于训练用于将表征与源语音相关联的源语音信号的源编码参数转换成表征与目标语音相关联的目标语音信号的相应目标编码参数的语音转换模型的转换器。 为了减少噪声对语音转换模型的影响，转换器可以被配置为用于接收源和目标编码参数的序列，并训练没有源和目标语音信号的能量小于阈值的一个或多个帧的模型 能源。 在转换各个参数之后，转换器，解码器或另一个组件可以被配置为用于减少具有小于阈值能量的能量的目标语音信号的一个或多个帧的能量，其中阈值可以是 基于语音帧和非语音帧的模型来适应。

公开(公告)号：US20230211135A1

公开(公告)日：2023-07-06

申请号：US18092270

申请日：2022-12-31

Applicant: Victor Popa-Simil

Inventor： Victor Popa-Simil

IPC: A61M27/00

CPC classification number: A61M27/006 ,  A61M2205/0238 ,  A61M2205/0244 ,  A61M2205/3584 ,  A61M2205/3344 ,  A61M2205/3368

Abstract: A method and device used to improve the operation of a hydrocephalus shunt system based on the use of alpha and beta radioactive isotopes implanted in the critical zones of the shunt in order to prevent the deposition of organic matter such as blood cells, tissue, or bacteria, thereby clogging the system and causing malfunction.

公开(公告)号：US10551826B2

公开(公告)日：2020-02-04

申请号：US15079387

申请日：2016-03-24

Applicant: Andrei Popa-Simil ,  Victor Popa-Simil

Inventor： Andrei Popa-Simil ,  Victor Popa-Simil

IPC: G06T19/00 ,  G05B19/418

Abstract: A method and system to increase operator awareness for process control application providing operators team real time information on controlled process and equipment immersing them into augmented and virtual reality, provided by a computing system which collects and integrates data from process and equipment, leaving equipment, controls, and procedure unmodified. The method is creating an augmented and virtual reality for the operating crew, giving them real time supplementary information based on two types of simulation, one using models and one using process data and learning procedures, together with rich equipment data, equipment locator, enhanced communication and enhanced data acquisition from supplementary sources as surface and airborne robotic and operator headset additional equipment meant to monitor operator bio-parameters and operator's surrounding environment in order to assure that control is sound and sober, and operators are safe and secure all time providing an optimum high efficiency operation of the controlled process.