公开(公告)号：WO2014144579A1

公开(公告)日：2014-09-18

申请号：PCT/US2014/029050

申请日：2014-03-14

Applicant: APPLE INC.

Inventor： TACKIN, Onur ,  NAIK, Devang, K.

IPC: G10L15/06 ,  G10L15/065 ,  G10L15/22 ,  G10L15/30 ,  G10L15/26

CPC classification number: G10L15/063 ,  G10L15/065 ,  G10L15/07 ,  G10L15/22 ,  G10L15/26 ,  G10L15/30 ,  G10L2015/223

Abstract: A method for updating an adaptive speech recognition model is provided. In some implementations, the method is performed at a communications device including one or more processors and memory storing instructions for execution by the one or more processors. The method includes determining that a first user of a first mobile communication device is engaged in a call over a communications network and providing an adaptive speech recognition model The method also includes tapping into an outbound audio channel of the first mobile communication device to obtain a call audio signal corresponding to audio input from one or more microphones of the first mobile communication device and updating the adaptive speech recognition model with training data derived from the call audio signal.

Abstract translation: 提供了一种用于更新自适应语音识别模型的方法。 在一些实施方式中，该方法在包括一个或多个处理器的通信设备和用于由一个或多个处理器执行的存储器存储指令的通信设备上执行。 该方法包括确定第一移动通信设备的第一用户通过通信网络参与呼叫并提供自适应语音识别模型。该方法还包括轻敲第一移动通信设备的出站音频信道以获得呼叫 对应于从第一移动通信设备的一个或多个麦克风输入的音频的音频信号，并使用从呼叫音频信号导出的训练数据来更新自适应语音识别模型。

公开(公告)号：WO2014124332A2

公开(公告)日：2014-08-14

申请号：PCT/US2014/015418

申请日：2014-02-07

Applicant: APPLE INC.

Inventor： BINDER, Justin, G. ,  TACKIN, Onur ,  POST, Samuel, D. ,  GRUBER, Thomas, R.

IPC: G10L17/24

CPC classification number: G10L21/16 ,  G06F3/167 ,  G10L15/22 ,  G10L15/265 ,  G10L15/30 ,  G10L17/24

Abstract: A method for operating a voice trigger is provided. In some implementations, the method is performed at an electronic device including one or more processors and memory storing instructions for execution by the one or more processors. The method includes receiving a sound input. The sound input may correspond to a spoken word or phrase, or a portion thereof. The method includes determining whether at least a portion of the sound input corresponds to a predetermined type of sound, such as a human voice. The method includes, upon a determination that at least a portion of the sound input corresponds to the predetermined type, determining whether the sound input includes predetermined content, such as a predetermined trigger word or phrase. The method also includes, upon a determination that the sound input includes the predetermined content, initiating a speech-based service, such as a voice-based digital assistant.

Abstract translation: 提供了一种用于操作语音触发器的方法。 在一些实现中，该方法在包括一个或多个处理器和用于存储由一个或多个处理器执行的指令的存储器的电子设备处执行。 该方法包括接收声音输入。 声音输入可以对应于说出的单词或短语或其一部分。 该方法包括确定声音输入的至少一部分是否对应于预定类型的声音，诸如人声。 该方法包括，在确定声音输入的至少一部分对应于预定类型时，确定声音输入是否包括预定内容，诸如预定的触发词或短语。 该方法还包括在确定声音输入包括预定内容时，发起基于语音的服务，诸如基于语音的数字助理。

公开(公告)号：WO2014197588A1

公开(公告)日：2014-12-11

申请号：PCT/US2014/040901

申请日：2014-06-04

Applicant: APPLE INC.

Inventor： VERGER, Luciano M. ,  CHAUDHARY, Madhusudan ,  TACKIN, Onur ,  ELRAD, Oren M. ,  VASUDEVAN, Srinivasan ,  ELANGOVAN, Thanigaivelu ,  SHAO, Tiange

IPC: H04J3/06 ,  H04W36/14 ,  H04L1/20 ,  H04L29/06

CPC classification number: H04W56/003 ,  H04J3/0632 ,  H04L1/203 ,  H04L1/205 ,  H04L29/06027 ,  H04L65/80 ,  H04L67/322 ,  H04W36/0022 ,  H04W36/0072 ,  H04W36/0088 ,  H04W36/08 ,  H04W36/14

Abstract: A jitter buffer in a Voice over LTE receiver may be influenced by radio level feedback (RLF) from both local and remote endpoints to preemptively adjust the jitter buffer delay in anticipation of predicted future losses that have a high probability of occurring. The radio events of the RLF and the scenarios that trigger the preemptive adjustments may be identified, and their use may be expressed in terms of mathematical formulas. In prior art designs, the instantaneous jitter is derived from a weighted history of the media stream, and consequently only packets that have already arrived are used to compute the instantaneous jitter to adjust the length of the buffer. By providing and using RLF from both local and remote endpoints, the anticipated delay - for packets that have not yet arrived - may be used to preemptively adjust the buffer, thereby minimizing packet loss without introducing unnecessary delay.

Abstract translation: 在LTE语音接收机中的抖动缓冲器可能受到来自本地和远端端点的无线电级别反馈（RLF）的影响，以预先调整抖动缓冲器延迟以预期具有高概率发生的预测未来损失。 可以识别RLF的无线电事件和触发抢先调整的场景，并且可以用数学公式来表示它们的使用。 在现有技术的设计中，瞬时抖动是从媒体流的加权历史导出的，因此只有已经到达的分组被用于计算瞬时抖动以调整缓冲器的长度。 通过从本地端点和远端端点提供和使用RLF，对于尚未到达的数据包来说，预期的延迟可能用于抢先调整缓冲区，从而最大限度地减少数据包丢失而不引入不必要的延迟。

公开(公告)号：WO2015048661A2

公开(公告)日：2015-04-02

申请号：PCT/US2014/058081

申请日：2014-09-29

Applicant: APPLE INC.

Inventor： VERGER, Luciano ,  THEVERAPPERUMA, Lalin ,  ALAM, Muhammad ,  DEIVASIGAMANI, Giri ,  KHAY-IBBAT, Samy ,  VASUDEVAN, Srinivasan ,  ELANGOVAN, Thanigaivelu ,  CHAUDHARY, Madhusudan ,  DAMJI, Navid ,  SEBENI, Johnson ,  SHI, Jianxiong ,  MAKHARIA, Shivesh ,  KODALI, Sree ,  TACKIN, Onur

CPC classification number: H04W52/0216 ,  G10L19/00 ,  H04L65/602 ,  H04W72/0413 ,  H04W72/1284 ,  H04W76/28 ,  Y02D70/1262 ,  Y02D70/23 ,  Y02D70/24

Abstract: A station that generates data packets to be transmitted such that the data packets spend a minimum amount of time in a buffer prior to transmission. The method includes receiving a specification for a connected discontinuous reception (C-DRX) cycle that indicates when a plurality of on Durations of the C-DRX cycle occurs, the on Durations having a predetermined interval therebetween, receiving data at a known time relative to the C-DRX cycle, determining a modification to a conversion process that converts the data to data packets such that the data packets are stored in a buffer at a subframe immediately preceding one of the on Durations subsequent to the known time, performing the conversion process based upon the modification and storing the data packets at the subframe immediately preceding the one of the on Durations. In one embodiment, the data is raw audio data and the data packets are audio packets.

Abstract translation: 生成要发送的数据分组的站，使得数据分组在传输之前在缓冲区中花费最少量的时间。 该方法包括：接收指示何时发生C-DRX周期的多个开启持续时间的连接的不连续接收（C-DRX）周期的规范，所述开启持续时间之间具有预定间隔;接收相对于 所述C-DRX周期确定对转换过程的修改，所述转换过程将所述数据转换为数据分组，使得所述数据分组被存储在紧随所述已知时间之后的所述持续时间之一之前的子帧处的缓冲器中，执行所述转换过程 基于修改并将数据分组存储在紧挨持续时间之一的子帧处。 在一个实施例中，数据是原始音频数据并且数据包是音频包。

公开(公告)号：EP3809407A1

公开(公告)日：2021-04-21

申请号：EP20198363.2

申请日：2014-02-07

Applicant: Apple Inc.

Inventor： BINDER, Justin G. ,  TACKIN, Onur ,  POST, Samuel D. ,  GRUBER, Thomas R.

IPC: G10L15/22 ,  G10L17/24 ,  G06F3/16 ,  G10L25/51 ,  G10L25/84

Abstract: A method for operating a voice trigger is provided. In some implementations, the method is performed at an electronic device including one or more processors and memory storing instructions for execution by the one or more processors. The method includes receiving a sound input. The sound input may correspond to a spoken word or phrase, or a portion thereof. The method includes determining whether at least a portion of the sound input corresponds to a predetermined type of sound, such as a human voice. The method includes, upon a determination that at least a portion of the sound input corresponds to the predetermined type, determining whether the sound input includes predetermined content, such as a predetermined trigger word or phrase. The method also includes, upon a determination that the sound input includes the predetermined content, initiating a speech-based service, such as a voice-based digital assistant.

公开(公告)号：EP2954514A2

公开(公告)日：2015-12-16

申请号：EP14707872.9

申请日：2014-02-07

Applicant: Apple Inc.

Inventor： BINDER, Justin, G. ,  TACKIN, Onur ,  POST, Samuel, D. ,  GRUBER, Thomas, R.

IPC: G10L17/24 ,  G06F3/16

CPC classification number: G10L21/16 ,  G06F3/167 ,  G10L15/22 ,  G10L15/265 ,  G10L15/30 ,  G10L17/24

Abstract: A method for operating a voice trigger is provided. In some implementations, the method is performed at an electronic device including one or more processors and memory storing instructions for execution by the one or more processors. The method includes receiving a sound input. The sound input may correspond to a spoken word or phrase, or a portion thereof. The method includes determining whether at least a portion of the sound input corresponds to a predetermined type of sound, such as a human voice. The method includes, upon a determination that at least a portion of the sound input corresponds to the predetermined type, determining whether the sound input includes predetermined content, such as a predetermined trigger word or phrase. The method also includes, upon a determination that the sound input includes the predetermined content, initiating a speech-based service, such as a voice-based digital assistant.