-
公开(公告)号：US20210366464A1
公开(公告)日：2021-11-25
申请号：US17397780
申请日：2021-08-09
Applicant: Intel Corporation
Inventor： Piotr Rozen , Joachim Hofer
Abstract: A system, article, and method of automatic speech recognition with highly efficient decoding is accomplished by frequent beam width adjustment.
-
公开(公告)号：US20210004686A1
公开(公告)日：2021-01-07
申请号：US17031496
申请日：2020-09-24
Applicant: Intel Corporation
Inventor： Piotr Rozen , Georg Stemmer
Abstract: Techniques related to implementing neural networks for speech recognition systems are discussed. Such techniques may include processing a node of the neural network by determining a score for the node as a product of weights and inputs such that the weights are fixed point integer values, applying a correction to the score based on a correction value associated with at least one of the weights, and generating an output from the node based on the corrected score.
-
公开(公告)号：US11031005B2
公开(公告)日：2021-06-08
申请号：US16223078
申请日：2018-12-17
Applicant: Intel Corporation
Inventor： Georg Stemmer , Andrzej Mialkowski , Joachim Hofer , Piotr Rozen , Tomasz Szmelczynski
Abstract: A mechanism is described for facilitating continuous topic detection and adaption in audio environments, according to one embodiment. A method of embodiments, as described herein, includes detecting a term relating to a topic in an audio input received from one or more microphones of the computing device including a voice-enabled device; analyzing the term based on the topic to determine an action to be performed by the computing device; and triggering an event to facilitate the computing device to perform the action consistent with the term and the topic.
-
公开(公告)号：US11735164B2
公开(公告)日：2023-08-22
申请号：US17397780
申请日：2021-08-09
Applicant: Intel Corporation
Inventor： Piotr Rozen , Joachim Hofer
Abstract: A system, article, and method of automatic speech recognition with highly efficient decoding is accomplished by frequent beam width adjustment.
-
5.发明授权Frame skipping with extrapolation and outputs on demand neural network for automatic speech recognition 有权帧跳跃与外推和输出按需神经网络进行自动语音识别
公开(公告)号：US09520128B2
公开(公告)日：2016-12-13
申请号：US14493434
申请日：2014-09-23
Applicant: Intel Corporation
Inventor： Josef Bauer , Piotr Rozen , Georg Stemmer
Abstract: Techniques related to implementing neural networks for speech recognition systems are discussed. Such techniques may include implementing frame skipping with approximated skip frames and/or distances on demand such that only those outputs needed by a speech decoder are provided via the neural network or approximation techniques.
Abstract translation: 讨论了与语音识别系统实现神经网络相关的技术。 这样的技术可以包括实现具有近似跳过帧和/或按需的距离的跳帧，使得仅通过神经网络或近似技术提供语音解码器所需的那些输出。
-
公开(公告)号：US10803381B2
公开(公告)日：2020-10-13
申请号：US15500405
申请日：2014-09-09
Applicant: INTEL CORPORATION
Inventor： Piotr Rozen , Georg Stemmer
IPC: G06N3/08 , G10L15/22 , G11C11/54 , A61B5/00 , G06N3/063 , G06F7/483 , G06N3/04 , G10L15/16 , G10L15/02
Abstract: Techniques related to implementing neural networks for speech recognition systems are discussed. Such techniques may include processing a node of the neural network by determining a score for the node as a product of weights and inputs such that the weights are fixed point integer values, applying a correction to the score based on a correction value associated with at least one of the weights, and generating an output from the node based on the corrected score.
-
公开(公告)号：US20200227024A1
公开(公告)日：2020-07-16
申请号：US16832410
申请日：2020-03-27
Applicant: Intel Corporation
Inventor： Piotr Rozen , Joachim Hofer
Abstract: A system, article, and method of automatic speech recognition with highly efficient decoding is accomplished by frequent beam width adjustment.
-
公开(公告)号：US20190087225A1
公开(公告)日：2019-03-21
申请号：US16191933
申请日：2018-11-15
Applicant: Intel Corporation
Inventor： Piotr Rozen , Sagar Koorapati
Abstract: Systems, apparatuses and methods may provide for technology that assigns a plurality of data portions associated with a workload to a plurality of cores, wherein each data portion from the plurality of data portions is only modifiable by a respective one of the plurality of cores. The technology may further pass a message between the plurality of cores to modify one or more of the data portions in response to an identification that the one or more of the data portions are unmodifiable by one or more of the plurality of cores.
-
公开(公告)号：US11120786B2
公开(公告)日：2021-09-14
申请号：US16832410
申请日：2020-03-27
Applicant: Intel Corporation
Inventor： Piotr Rozen , Joachim Hofer
Abstract: A system, article, and method of automatic speech recognition with highly efficient decoding is accomplished by frequent beam width adjustment.
-
公开(公告)号：US10650807B2
公开(公告)日：2020-05-12
申请号：US16134826
申请日：2018-09-18
Applicant: Intel Corporation
Inventor： Tobias Bocklet , Jacek Ossowski , Tomasz Dorau , Maciej Muchlinski , David Pearce , Piotr Rozen
IPC: G10L15/00 , G10L15/16 , G10L15/22 , G06N7/00 , G06N3/063 , G10L15/02 , G06N3/04 , G10L15/08 , G10L25/84 , G10L25/30
Abstract: A method and system are directed to autonomous neural network keyphrase detection and includes generating and using a multiple element state score vector by using neural network operations and without substantial use of a digital signal processor (DSP) to perform the keyphrase detection.
