公开(公告)号：US20120224072A1

公开(公告)日：2012-09-06

申请号：US13039875

申请日：2011-03-03

申请人： Hyung-II Koo ,  Taesu Kim ,  Ki-Hyun Kim ,  Te-Won Lee

发明人： Hyung-II Koo ,  Taesu Kim ,  Ki-Hyun Kim ,  Te-Won Lee

IPC分类号： H04N5/228

CPC分类号： G06K9/22 ,  G06K9/325 ,  G06K2209/01 ,  H04N5/23267 ,  H04N5/2327

摘要： Techniques are described for identifying blurred images and recognizing text. One or more images of text may be captured. A change of movement associated with each image of the one or more images may be calculated. The change of movement associated with an image of the one or more images represents a change in an amount of acceleration of the device used to capture the image while the image was being captured. A steady image may be selected from the one or more images to use for text recognition. The steady image can be selected using the variances of acceleration associated with each image of the one or more images.

摘要翻译： 描述了识别模糊图像和识别文本的技术。 可以捕获一个或多个文本图像。 可以计算与一个或多个图像的每个图像相关联的移动变化。 与一个或多个图像的图像相关联的运动的变化表示在拍摄图像时用于捕获图像的装置的加速度的变化。 可以从用于文本识别的一个或多个图像中选择稳定的图像。 可以使用与一个或多个图像的每个图像相关联的加速度的方差来选择稳定图像。

公开(公告)号：US20090222262A1

公开(公告)日：2009-09-03

申请号：US12281298

申请日：2006-03-01

申请人： Taesu Kim ,  Te-Won Lee

发明人： Taesu Kim ,  Te-Won Lee

IPC分类号： G10L15/02 ,  G10L11/00

CPC分类号： G10L21/0272 ,  G10L25/84

摘要： Signal separation techniques based on frequency dependency are described. In one implementation, a blind signal separation process is provided that avoids the permutation problem of previous signal separation processes. In the process, two or more signal sources are provided, with each signal source having recognized frequency dependencies. The process uses these inter-frequency dependencies to more robustly separate the source signals. The process receives a set of mixed signal input signals, and samples each input signal using a rolling window process. The sampled data is transformed into the frequency domain, which provides channel inputs to the inter-frequency dependent separation process. Since frequency dependencies have been defined for each source, the process is able to use the frequency dependency to more accurately separate the signals. The process can use a learning algorithm that preserves frequency dependencies within each source signal, and can remove dependencies between or among the signal sources.

摘要翻译： 描述了基于频率依赖性的信号分离技术。 在一个实现中，提供了一种盲信号分离过程，其避免了先前信号分离过程的置换问题。 在该过程中，提供两个或更多个信号源，每个信号源具有已识别的频率依赖性。 该过程使用这些频率间依赖性来更鲁棒地分离源信号。 该过程接收一组混合信号输入信号，并使用滚动窗口进程对每个输入信号进行采样。 采样数据被转换成频域，其向频率间依赖分离过程提供信道输入。 由于为每个源定义了频率相关性，因此该过程能够使用频率依赖性来更准确地分离信号。 该过程可以使用保持每个源信号内的频率依赖性的学习算法，并且可以消除信号源之间或之间的依赖关系。

公开(公告)号：US09143571B2

公开(公告)日：2015-09-22

申请号：US13370668

申请日：2012-02-10

申请人： Taesu Kim ,  Te-Won Lee

发明人： Taesu Kim ,  Te-Won Lee

IPC分类号： G06F17/00 ,  H04L29/08 ,  H04W4/02 ,  H04W4/20 ,  H04W4/04 ,  H04W4/08

CPC分类号： H04L67/18 ,  H04W4/023 ,  H04W4/043 ,  H04W4/08 ,  H04W4/21

摘要： A method for identifying mobile devices in a similar sound environment is disclosed. Each of at least two mobile devices captures an input sound and extracts a sound signature from the input sound. Further, the mobile device extracts a sound feature from the input sound and determines a reliability value based on the sound feature. The reliability value may refer to a probability of a normal sound class given the sound feature. A server receives a packet including the sound signatures and reliability values from the mobile devices. A similarity value between sound signatures from a pair of the mobile devices is determined based on corresponding reliability values from the pair of mobile devices. Specifically, the sound signatures are weighted by the corresponding reliability values. The server identifies mobile devices in a similar sound environment based on the similarity values.

摘要翻译： 公开了一种用于识别类似声音环境中的移动设备的方法。 至少两个移动设备中的每一个捕获输入声音并从输入声音提取声音签名。 此外，移动设备从输入声音提取声音特征，并且基于声音特征确定可靠性值。 可靠性值可以指给定声音特征的正常声级的概率。 服务器从移动设备接收包括声音签名和可靠性值的数据包。 来自一对移动设备的声音签名之间的相似性值是基于来自该对移动设备的相应的可靠性值确定的。 具体地说，声音特征被相应的可靠性值加权。 服务器基于相似性值识别类似声音环境中的移动设备。

公开(公告)号：US08874439B2

公开(公告)日：2014-10-28

申请号：US12281298

申请日：2006-03-01

申请人： Taesu Kim ,  Te-Won Lee

发明人： Taesu Kim ,  Te-Won Lee

IPC分类号： G10L21/02 ,  G10L21/0272 ,  G10L25/84

CPC分类号： G10L21/0272 ,  G10L25/84

摘要： Signal separation techniques based on frequency dependency are described. In one implementation, a blind signal separation process is provided that avoids the permutation problem of previous signal separation processes. In the process, two or more signal sources are provided, with each signal source having recognized frequency dependencies. The process uses these inter-frequency dependencies to more robustly separate the source signals. The process receives a set of mixed signal input signals, and samples each input signal using a rolling window process. The sampled data is transformed into the frequency domain, which provides channel inputs to the inter-frequency dependent separation process. Since frequency dependencies have been defined for each source, the process is able to use the frequency dependency to more accurately separate the signals. The process can use a learning algorithm that preserves frequency dependencies within each source signal, and can remove dependencies between or among the signal sources.

摘要翻译： 描述了基于频率依赖性的信号分离技术。 在一个实现中，提供了一种盲信号分离过程，其避免了先前信号分离过程的置换问题。 在该过程中，提供两个或更多个信号源，每个信号源具有已识别的频率依赖性。 该过程使用这些频率间依赖性来更鲁棒地分离源信号。 该过程接收一组混合信号输入信号，并使用滚动窗口进程对每个输入信号进行采样。 采样数据被转换成频域，其向频率间依赖分离过程提供信道输入。 由于为每个源定义了频率相关性，因此该过程能够使用频率依赖性来更准确地分离信号。 该过程可以使用保持每个源信号内的频率依赖性的学习算法，并且可以消除信号源之间或之间的依赖关系。

公开(公告)号：US20130258141A1

公开(公告)日：2013-10-03

申请号：US13550531

申请日：2012-07-16

申请人： Taesu Kim ,  Hyun-Mook Cho ,  Hyung-Il Koo

发明人： Taesu Kim ,  Hyun-Mook Cho ,  Hyung-Il Koo

IPC分类号： H04N5/235

CPC分类号： H04N5/235 ,  G06T7/246 ,  G06T2207/10016 ,  G06T2207/30244

摘要： Methods, apparatuses, systems, and computer-readable media for rejecting false positive detection and tracking of image objects are presented. According to one or more aspects, a computing device may implement embodiments of the invention to use the movement of the mobile device for distinguishing false positives from true movement of the mobile device depicted in the field of view of the camera. In one embodiment, the actual movement of the mobile device may be measured using multi-modal sensor data from inertial sensors such as accelerometers and gyroscopes. In another embodiment, the actual movement of the device is calculated using the global movement of the mobile phone with reference to other objects in the field of view of the camera.

摘要翻译： 提出了用于拒绝图像对象的假阳性检测和跟踪的方法，装置，系统和计算机可读介质。 根据一个或多个方面，计算设备可以实现本发明的实施例，以使用移动设备的移动来区分在相机视野中描绘的移动设备的真实移动的假阳性。 在一个实施例中，可以使用来自诸如加速度计和陀螺仪的惯性传感器的多模态传感器数据来测量移动设备的实际移动。 在另一个实施例中，使用移动电话参照相机视野中的其他对象的全局移动来计算设备的实际移动。

公开(公告)号：US20110013075A1

公开(公告)日：2011-01-20

申请号：US12839338

申请日：2010-07-19

申请人： Taesu Kim ,  Seungil Kim ,  Hyejeong Jeon ,  Gunho Song ,  Sangwhan Moon

发明人： Taesu Kim ,  Seungil Kim ,  Hyejeong Jeon ,  Gunho Song ,  Sangwhan Moon

IPC分类号： H04R3/00 ,  H04N5/225

CPC分类号： H04N5/602 ,  H04M1/6008 ,  H04N7/147 ,  H04N21/41407 ,  H04N21/4788 ,  H04R3/005 ,  H04R2420/01 ,  H04R2430/03 ,  H04R2499/11

摘要： The present disclosure relates to a method for processing sound source in terminal capable of receiving an outside sound source through at least two microphones and displaying a position of the outside sound source on a display unit, and a terminal using the same, wherein the terminal includes: at least two microphones configured to receive an outside sound source; a display unit configured to display a predetermined data; and a controller configured to obtain position information of the sound source using at least one of an amplitude, a phase and a period of the outside sound source received from the two microphones, and controllably display the position information of the sound source on a display unit.

摘要翻译： 本公开涉及一种用于处理终端中的声源的方法，其能够通过至少两个麦克风接收外部声源并且在显示单元上显示外部声源的位置，以及使用其的终端，其中，所述终端包括 ：配置成接收外部声源的至少两个麦克风; 配置为显示预定数据的显示单元; 以及控制器，被配置为使用从两个麦克风接收的外部声源的幅度，相位和周期中的至少一个获得声源的位置信息，并且可控地在显示单元上显示声源的位置信息 。