专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明专利

JP2011182142A Sound signal false localization system, method thereof, sound signal false localization decoding device and program 有权
标题翻译：声信号伪定位系统及其方法，声信号伪定位解码器和程序
公开(公告)号：JP2011182142A
公开(公告)日：2011-09-15
申请号：JP2010043457
申请日：2010-02-26
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： TSUTSUMI KIMITAKA , SASAKI SHIGEAKI , OMURO NAKA , HIWAZAKI YUUSUKE , FUKUI KATSUHIRO
IPC分类号： H04S1/00 , G10L19/00 , H04M3/56 , H04N7/15 , H04S5/02
摘要： PROBLEM TO BE SOLVED: To add sound signals from multiple points without impairing the presence in a multi-point communication conference system etc. SOLUTION: In this system, an MS converting section 3 adds a first input signal to a second input signal to form an M signal, and subtracts the second input signal from the first input signal so as to form an S signal. An expanded false localization imparting section 1 calculates a decoded M signal S o M (t) and a decoded S signal S o S (t) defined by S o M (t)=S M (t), S o S (t)=(N-2n-1)*S M (t)/N+(1/N)*S S (t), where a point number N≥2 is the total number of directions where localization can be imparted, point information n is the number of a direction where localization is imparted, the M signal is S M (t) and the S signal is S S (t). An LR converting section 2 generates the first output signal by adding the decoded M signal to the decoded S signal and dividing the sum by 2, and generates the second output signal by subtracting the decoded S signal from the decoded M signal and dividing the balance by 2. COPYRIGHT: (C)2011,JPO&INPIT
摘要翻译：要解决的问题：从多点添加声音信号，而不损害在多点通信会议系统中的存在等。解决方案：在该系统中，MS转换部分3将第一输入信号添加到第二输入信号以形成M信号，并从第一输入信号中减去第二输入信号以形成S信号。扩展的假定位赋予部分1计算解码的M信号S（SP）和解码的S信号S o S （t）由S o定义的（t）（t）=（S） SP> 取值（T）=（N-2 N-1）* S 中号（T）/ N +（1 / N）* S 取值（t），其中点数N≥2是可以赋予定位的方向的总数，点信息n是赋予定位的方向的数量，M信号是S M （t），S信号为S（S）。 LR转换部分2通过将解码的M信号与解码的S信号相加并将和除以2产生第一输出信号，并且通过从解码的M信号中减去解码的S信号并将其余数除以 2.版权所有（C）2011，JPO＆INPIT

2. 发明专利

JP2011182141A Sound signal false localization system, method and program 有权
标题翻译：声信号伪定位系统，方法与程序
公开(公告)号：JP2011182141A
公开(公告)日：2011-09-15
申请号：JP2010043456
申请日：2010-02-26
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： TSUTSUMI KIMITAKA , SASAKI SHIGEAKI , OMURO NAKA , HIWAZAKI YUUSUKE , FUKUI KATSUHIRO
IPC分类号： H04S1/00 , G10L19/00 , H04R3/00 , H04S5/00
摘要： PROBLEM TO BE SOLVED: To reduce the amount of processing of a mixer in a sound signal false localization system. SOLUTION: A gain decoding section 8 of the mixer 300 decodes only the gain, and a false localization gain operation section 9 generates a false M gain and a false S gain which are subjected to false localization by using the decoded gain (decoded M gain). The mixer 300 imparts false localization by using the false M gain and the false S gain which are generated. As described above, the mixer decodes not the entire codes, but only the gain to impart the false localization by using the decoded gain. Thus, the amount of processing of mixer becomes smaller than in a conventional system. COPYRIGHT: (C)2011,JPO&INPIT
摘要翻译：要解决的问题：减少声信号虚拟定位系统中混频器的处理量。解决方案：混合器300的增益解码部分8仅对增益进行解码，并且假定位增益操作部分9通过使用解码的增益（解码的）产生假M定位和假S增益 M增益）。混合器300通过使用产生的假M增益和假S增益来赋予虚拟定位。如上所述，混频器不对整个码进行解码，而只解码通过使用解码增益来赋予假定位的增益。因此，混合器的处理量比常规系统中的小。版权所有（C）2011，JPO＆INPIT

3. 发明专利

JP2011118408A Coding method and program 有权
标题翻译：编码方法和程序
公开(公告)号：JP2011118408A
公开(公告)日：2011-06-16
申请号：JP2011007985
申请日：2011-01-18
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： OMURO NAKA , SASAKI SHIGEAKI , HIWAZAKI YUUSUKE , MORI TAKESHI , KATAOKA AKITOSHI
IPC分类号： G10L19/00 , H03M7/36
摘要： PROBLEM TO BE SOLVED: To provide a coding device that facilitates combination with an existing coding system, and effectively reduces a harsh feeling of quantization noise only by adding a function only to a transmitting side.
SOLUTION: A method includes: a step of obtaining addition results by adding a digital sound signal sample x
n of the n-th sample and a feedback signal sample; a step of generating a coding signal from the addition results by a predetermined coding system for coding an input signal at every sample; a step of generating a decoding signal sample by a decoding system corresponding to the coding system from the coding signal; a step of obtaining subtraction results by subtracting the digital sound signal sample from the decoding signal sample; and a delay step of obtaining the feedback signal sample by delaying the subtraction results by one sampling time. The steps are repeatedly executed while the input of the digital sound signal sample x
n is continued as x
n+1 , x
n+2 .
COPYRIGHT: (C)2011,JPO&INPIT
摘要翻译：要解决的问题：提供一种便于与现有编码系统组合的编码装置，并且仅通过向发送侧添加功能来有效地降低量化噪声的粗糙感。解决方案：一种方法包括：通过添加第n个样本和反馈信号样本的数字声音信号样本x n 获得加法结果的步骤; 通过用于对每个采样的输入信号进行编码的预定编码系统从加法结果生成编码信号的步骤; 通过与编码系统对应的解码系统从编码信号生成解码信号样本的步骤; 通过从解码信号样本中减去数字声音信号样本来获得减法结果的步骤; 以及通过将减法结果延迟一个采样时间来获得反馈信号采样的延迟步骤。当数字声音信号样本x n 的输入继续进行为x n + 1 ，x n + 2 时，重复执行这些步骤。版权所有（C）2011，JPO＆INPIT

4. 发明专利

JP2011009869A Encoding method, decoding method, encoder, decoder, and program 有权
标题翻译：编码方法，解码方法，编码器，解码器和程序
公开(公告)号：JP2011009869A
公开(公告)日：2011-01-13
申请号：JP2009149008
申请日：2009-06-23
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： FUKUI KATSUHIRO , SASAKI SHIGEAKI , TSUTSUMI KIMITAKA , HIWAZAKI YUUSUKE
IPC分类号： H03M7/30
摘要： PROBLEM TO BE SOLVED: To reduce musical noise suitably in accordance with a type of first signals to be encoded even when a sufficient code bit length can not be secured.SOLUTION: A vector comprising a predetermined number of first signals to be encoded or a predetermined number of signals corresponding thereto as elements is quantized, and first quantization signals corresponding to the first signals to be encoded and first quantization indexes for specifying the quantization signals are generated and outputted. On the basis of a decision value specifying the number of first signals to be encoded which have amplitude values smaller than reference value or the number of first signals to be encoded whose amplitude values are equal to or smaller than the reference value among the predetermined number of first signals to be encoded, an encoding system corresponding to the predetermined number of first signals to be encoded is selected, and mode information specifying the selected encoding system is generated and output to performing encoding processing at least using the first quantization signals according to the selected encoding system.
摘要翻译：要解决的问题：即使当无法确保足够的码位长度时，也可以根据要编码的第一信号的类型来适当地减小音乐噪声。解决方案：包括预定数量的要编码的第一信号或预定的量化对应于其的信号数量，生成并输出用于指定量化信号的与要编码的第一信号对应的第一量化信号和用于指定量化信号的第一量化索引。基于指定要编码的第一信号的数量的决定值，该第一信号的幅度值小于参考值，或者在预定数量的参数值中幅度值等于或小于参考值的要编码的第一信号的数量选择要编码的第一信号，选择与要编码的预定数量的第一信号相对应的编码系统，并且生成指定所选编码系统的模式信息，并输出至少使用根据所选择的第一量化信号进行编码处理编码系统。

5. 发明专利

JP2011007870A Encoding method, decoding method, encoding device, decoding device, encoding program and decoding program 有权
标题翻译：编码方法，解码方法，编码设备，解码设备，编码程序和解码程序
公开(公告)号：JP2011007870A
公开(公告)日：2011-01-13
申请号：JP2009148887
申请日：2009-06-23
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： TSUTSUMI KIMITAKA , SASAKI SHIGEAKI , HIWAZAKI YUUSUKE , FUKUI KATSUHIRO
IPC分类号： G10L19/00 , G10L19/02
摘要： PROBLEM TO BE SOLVED: To provide technology capable of reducing a storage area required for encoding and decoding.SOLUTION: A sound signal is constituted for each frame by collecting samples for constituting input sound signal with the predetermined number. Either a time domain encoding method or a frequency domain encoding method, by which a sound signal with higher quality is obtained in the case of decoding is selected, and signal identification information as a result of the selection result is output. When the time domain encoding method is selected, the sound signal of each frame is encoded with vector quantization by using a first code book without converting a time domain signal, and signal quantization code index which is a result of the encoding is output. When the frequency domain encoding method is selected, the sound signal of each frame is converted to the frequency domain signal, and it is encoded with vector quantization by using the first code book, and the signal quantization code index which is a result of the encoding is output.
摘要翻译：要解决的问题：提供能够减少编码和解码所需的存储区域的技术。解决方案：通过收集用于构成具有预定数量的输入声音信号的样本，为每个帧构成声音信号。选择在解码的情况下获得具有较高质量的声音信号的时域编码方法或频域编码方法，并且输出作为选择结果的结果的信号识别信息。当选择时域编码方法时，通过使用不转换时域信号的第一代码簿对每帧的声音信号进行矢量量化编码，并且输出作为编码结果的信号量化代码索引。当选择频域编码方法时，将每帧的声音信号转换为频域信号，并通过使用第一码本对矢量进行编码，作为编码结果的信号量化码索引被输出。

6. 发明专利

JP2008311752A Encoder, encoding method, program, and recording medium 有权
标题翻译：编码器，编码方法，程序和记录介质
公开(公告)号：JP2008311752A
公开(公告)日：2008-12-25
申请号：JP2007155413
申请日：2007-06-12
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： OMURO NAKA , SASAKI SHIGEAKI , HIWAZAKI YUUSUKE , MORI TAKESHI , KATAOKA AKITOSHI
IPC分类号： H03M7/30 , G10L19/00 , G10L19/14
摘要： PROBLEM TO BE SOLVED: To provide an encoder which can easily be combined with an existing encoding system and can effectively reduce a sense of quantization noise offensive to the ear simply by adding a function only to a transmission side. SOLUTION: An addition part 11 adds up a digital signal and a feedback signal input to the device and outputs an addition result. An encoder 101 encodes the addition result by a prescribed encoding system and outputs it. A decoder 12 decodes the encoded signal by a decoding system corresponding to the encoding system and outputs it. A subtraction part 13 subtracts the digital sound signal from the decoded signal and outputs a subtraction result. A delay part 14 delays the subtraction result for 1 sampling time and outputs the signal towards the addition part 11 as the feedback signal. COPYRIGHT: (C)2009,JPO&INPIT
摘要翻译：要解决的问题：提供一种可以容易地与现有编码系统组合的编码器，并且可以通过仅将功能添加到发送侧来有效地减少对耳朵攻击的量化噪声的感觉。解决方案：加法部分11将数字信号和反馈信号输入加到装置上并输出相加结果。编码器101通过规定的编码系统对相加结果进行编码并输出。解码器12通过与编码系统对应的解码系统对编码信号进行解码并输出。减法部13从解码信号中减去数字声音信号，并输出减法结果。延迟部分14将减法结果延迟1个采样时间，并将信号作为反馈信号输出到加法部分11。版权所有（C）2009，JPO＆INPIT

7. 发明专利

JP2012088502A Encoding device, decoding device, encoding method, decoding method, and program thereof 有权
标题翻译：编码设备，解码设备，编码方法，解码方法及其程序
公开(公告)号：JP2012088502A
公开(公告)日：2012-05-10
申请号：JP2010234728
申请日：2010-10-19
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： KAMAMOTO MASARU , HARADA NOBORU , MORIYA TAKEHIRO , OMURO NAKA , SASAKI SHIGEAKI , HIWAZAKI YUUSUKE
IPC分类号： G10L19/00 , H03M7/30
摘要： PROBLEM TO BE SOLVED: To provide an encoding technique capable of transmitting more extension layer codes in scalable encoding by effectively utilizing a bit reduction effect by lossless encoding.SOLUTION: In encoding according to the present invention, a digital input signal is scalably encoded to generate a basic layer code and one or more extension layer codes, the basic layer code is losslessly encoded to generate a lossless code, and a combination in which a code amount per unit time is equal to or less than a utilization transmission band and the code amount per unit time is maximum is selected from among plural kinds of combinations of the lossless code and the extension layer codes.
摘要翻译：要解决的问题：提供一种能够通过有效利用无损编码的比特减少效果来在可分级编码中发送更多扩展层编码的编码技术。解决方案：在根据本发明的编码中，数字输入信号被可缩放地编码以产生基本层代码和一个或多个扩展层代码，基本层代码被无损编码以产生无损代码，以及组合从无损代码和扩展层代码的多种组合中选择每单位时间的代码量等于或小于使用传输频带，并且每单位时间的代码量是最大的。版权所有（C）2012，JPO＆INPIT

8. 发明专利

JP2010066335A Signal broadband forming device, signal broadband forming method, program thereof and recording medium thereof 有权
标题翻译：信号宽带形成装置，信号宽带形成方法，程序及其记录介质
公开(公告)号：JP2010066335A
公开(公告)日：2010-03-25
申请号：JP2008230426
申请日：2008-09-09
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： MORI TAKESHI , SASAKI SHIGEAKI , OMURO NAKA , HIWAZAKI YUUSUKE , HANEDA YOICHI
IPC分类号： G10L21/04 , G10L11/00
摘要： PROBLEM TO BE SOLVED: To create a pseudo-broadband signal with good quality from a narrow band signal.
SOLUTION: A low frequency signal is created by converting the narrow band signal into a frequency domain. A PARCOR (PARtial auto-CORrelation ) coefficient is calculated by performing PARCOR analysis on the narrow band signal. A gain coefficient in which a value of a low frequency side becomes larger than that of a high frequency side, as the input PARCOR coefficient becomes smaller. An emphasized low frequency signal is created by multiplying the gain coefficient by the low frequency signal. A high frequency signal is created by copying a part or a whole of the low frequency signal. A pseudo-broadband frequency signal is created by arranging the emphasized low frequency signal at the low frequency side and the high frequency signal at the high frequency side and combining them. The pseudo-broadband signal is output by converting the pseudo-broadband frequency signal into a time domain.
COPYRIGHT: (C)2010,JPO&INPIT
摘要翻译：要解决的问题：从窄带信号创建质量好的伪宽带信号。解决方案：通过将窄带信号转换为频域来产生低频信号。通过对窄带信号执行PARCOR分析来计算PARCOR（PARtial auto-Correlation）系数。随着输入PARCOR系数变小，低频侧的值变得大于高频侧的增益系数。通过将增益系数乘以低频信号来产生强调的低频信号。通过复制低频信号的一部分或全部来产生高频信号。通过在低频侧布置强调的低频信号和高频侧的高频信号并组合来产生伪宽带频率信号。通过将伪宽带频率信号转换成时域来输出伪宽带信号。版权所有（C）2010，JPO＆INPIT

9. 发明专利

JP2009134260A Voice musical sound false broadband forming device, voice speech musical sound false broadband forming method, and its program and its record medium 有权
标题翻译：语音音乐宽带宽带形成设备，语音音乐声音无线宽带形成方法及其程序及其记录介质
公开(公告)号：JP2009134260A
公开(公告)日：2009-06-18
申请号：JP2008230455
申请日：2008-09-09
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： MORI TAKESHI , SASAKI SHIGEAKI , TSUTSUMI KIMITAKA , HIWAZAKI YUUSUKE , OMURO NAKA , KATAOKA AKITOSHI
IPC分类号： G10L21/04
摘要： PROBLEM TO BE SOLVED: To improve the voice quality of a voice musical sound false broadband forming device.
SOLUTION: The voice musical sound false broadband forming device converts a narrow band voice musical sound signal to a signal of a frequency region to form a signal of a low frequency region, multiplying the signal of the low frequency region by a gain coefficient to create a signal of a high frequency region, and synthesizes the signal of the high frequency region with the signal of the low frequency region to be a false broadband region. A gain determination section determines a gain so as to make the gain smaller when the power of the signal on a low-pass side or the absolute value of amplitude is large, and to make the gain larger in the case where the power of the signal on a high-pass side or the absolute value of amplitude is large on the basis of the power ratio or the absolute value of amplitude of the signal of different ranges within the signal on a low-frequency region.
COPYRIGHT: (C)2009,JPO&INPIT
摘要翻译：要解决的问题：提高语音音乐假宽带形成装置的语音质量。解决方案：声音声音假宽带形成装置将窄带声音音乐声音信号转换为频率区域的信号，以形成低频区域的信号，将低频区域的信号乘以增益系数以产生高频区域的信号，并将高频区域的信号与低频区域的信号合成为假宽带区域。增益确定部分确定增益，以便当低通侧的信号的功率或振幅的绝对值较大时使增益更小，并且在信号的功率的情况下使增益更大基于低频区域中的信号内的不同范围的信号的功率比或振幅的绝对值，在高通侧或幅度的绝对值较大。版权所有（C）2009，JPO＆INPIT

10. 发明专利

JP2009003388A Device for determination of pitch search range, pitch search device, packet loss compensation device, their methods, program and its recording medium 有权
标题翻译：用于确定搜索范围的设备，PITCH搜索设备，分组丢失补偿设备，其方法，程序及其记录介质
公开(公告)号：JP2009003388A
公开(公告)日：2009-01-08
申请号：JP2007166883
申请日：2007-06-25
申请人： Nippon Telegr & Teleph Corp , 日本電信電話株式会社
发明人： OMURO NAKA , SASAKI SHIGEAKI , HIWAZAKI YUUSUKE , MORI TAKESHI , KATAOKA AKITOSHI
IPC分类号： G10L19/00 , G10L11/04
摘要： PROBLEM TO BE SOLVED: To provide technology for determining a search range used in a pitch estimation method, which is used for packet loss concealment (packet loss compensation) etc.
SOLUTION: An adoption index calculation part 11 calculates an adoption index which is the index for expressing a vocality of an audio signal of a part in which the pitch is to be obtained. A judgment section 2 judges whether or not, the vocality of the audio signal in which the pitch is to be obtained is stronger than prescribed strength, by comparing the adoption index with a predetermined adoption reference value. When it is judged that the vocality of the audio signal of the part in which the pitch is to be obtained is stronger than the prescribed strength, a search range determination section 13 determines that a search range A is the search range, and otherwise, a search range B which is narrower than the search range A is the search range.
COPYRIGHT: (C)2009,JPO&INPIT
摘要翻译：要解决的问题：提供用于确定用于分组丢失隐藏（分组丢失补偿）等的音调估计方法中使用的搜索范围的技术。解决方案：采用指数计算部分11计算采用指数，其是表示要获得音调的部分的音频信号的声音的索引。判断部分2通过将采用指数与预定的采用参考值相比较来判断要获得音调的音频信号的声音是否比规定强度更强。当判断要获得音调的部分的音频信号的声音比规定强度更强时，搜索范围确定部分13确定搜索范围A是搜索范围，否则，搜索范围B比搜索范围A窄，是搜索范围。版权所有（C）2009，JPO＆INPIT

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式