公开(公告)号：US12205039B1

公开(公告)日：2025-01-21

申请号：US17087181

申请日：2020-11-02

Applicant: Amazon Technologies, Inc.

Inventor： Ritwik Giri ,  Srikanth Venkata Tenneti ,  Karim Helwani ,  Fangzhou Cheng ,  Mehmet Umut Isik ,  Arvindh Krishnaswamy

IPC: G06N3/088 ,  G10L25/03 ,  G10L25/51

Abstract: A group masked autoencoder may be implemented for anomaly detection. An autoencoder network model may be trained without supervision and applied to output an estimated joint probability distribution of normality for a group of frames of time series data. The estimated joint probability distribution may be used to determine an anomaly score for the time series data. An anomaly may be detected according to the anomaly score and a result that indicates a detected anomaly may be provided.

公开(公告)号：US11521637B1

公开(公告)日：2022-12-06

申请号：US17037498

申请日：2020-09-29

Applicant: Amazon Technologies, Inc.

Inventor： Jean-Marc Valin ,  Mehmet Umut Isik ,  Neerad Dilip Phansalkar ,  Ritwik Giri ,  Karim Helwani ,  Arvindh Krishnaswamy

IPC: G10L21/034 ,  G06F3/16 ,  G10L25/30

Abstract: Post-filtering may be performed for ratio masks as part of audio enhancement. Audio data may be received. A machine learning model may be applied to generate gain values for different spectrum bands of the audio data. The gain values may then be modified using an envelope post-filter according to a monotonically increasing function applied to the gain values to produce modified gain values used to generate an enhanced version of the audio data.

公开(公告)号：US12014748B1

公开(公告)日：2024-06-18

申请号：US16988423

申请日：2020-08-07

Applicant: Amazon Technologies, Inc.

Inventor： Ritwik Giri ,  Mehmet Umut Isik ,  Neerad Dilip Phansalkar ,  Jean-Marc Valin ,  Karim Helwani ,  Arvindh Krishnaswamy

IPC: G10L21/0208 ,  G06N5/04 ,  G06N20/00 ,  G10L21/034

CPC classification number: G10L21/0208 ,  G06N5/04 ,  G06N20/00 ,  G10L21/034 ,  G10L2021/02082

Abstract: Techniques for training and using a machine learning model for estimation of reverberation in a multi-task learning framework are described. According to some embodiments, the multi-task learning framework improves the performance of the machine learning model by estimating the amount of reverberation present in an input audio recording as a secondary task to the primary task of generating a clean speech portion of the input audio recording. In one embodiment, a model architecture is selected that takes a noisy reverberant recording as an input and outputs an estimate of a clean (e.g., de-reverberated) signal, an estimate of noise (e.g., background noise), and an estimate of the reverb only portion, with the secondary task of estimating the reverb only portion acting as a regularizer that improves the machine learning model's performance in enhancing the reverberant (e.g., and noisy) input speech.

公开(公告)号：US20240096346A1

公开(公告)日：2024-03-21

申请号：US17850617

申请日：2022-06-27

Applicant: Amazon Technologies, Inc.

Inventor： Masahito Togami ,  Ritwik Giri ,  Michael Mark Goodwin ,  Arvindh . Krishnaswamy ,  Siddhartha Shankara Rao

IPC: G10L21/10 ,  G10L15/04 ,  G10L21/0208

CPC classification number: G10L21/10 ,  G10L15/04 ,  G10L21/0208

Abstract: A plurality of talker embedding vectors may be derived that correspond to a plurality of talkers in an input audio stream. Each talker embedding vector may represent respective voice characteristics of a respective talker. The talker embedding vectors may be generated based on, for example, a pre-enrollment process or a cluster-based embedding vector derivation process. A plurality of instances of a personalized noise suppression model may be executed on the input audio stream. Each instance of the personalized noise suppression model may employ a respective talker embedding vector. A plurality of single-talker audio streams may be generated by the plurality of instances of the personalized noise suppression model. A plurality of single-talker transcriptions may be generated based on the plurality of single-talker audio streams. The plurality of single-talker transcriptions may be merged into a multi-talker output transcription.

公开(公告)号：US11545134B1

公开(公告)日：2023-01-03

申请号：US16709792

申请日：2019-12-10

Applicant: Amazon Technologies, Inc.

Inventor： Marcello Federico ,  Robert Enyedi ,  Yaser Al-Onaizan ,  Roberto Barra-Chicote ,  Andrew Paul Breen ,  Ritwik Giri ,  Mehmet Umut Isik ,  Arvindh Krishnaswamy ,  Hassan Sawaf

IPC: G10L13/08 ,  G10L15/22 ,  G11B20/10 ,  G06F3/16 ,  G10L13/10 ,  G06F40/47 ,  G10L25/90 ,  G10L15/06 ,  G10L13/00 ,  G10L15/26 ,  G06V40/16

Abstract: Techniques for the generation of dubbed audio for an audio/video are described. An exemplary approach is to receive a request to generate dubbed speech for an audio/visual file; and in response to the request to: extract speech segments from an audio track of the audio/visual file associated with identified speakers; translate the extracted speech segments into a target language; determine a machine learning model per identified speaker, the trained machine learning models to be used to generate a spoken version of the translated, extracted speech segments based on the identified speaker; generate, per translated, extracted speech segment, a spoken version of the translated, extracted speech segments using a trained machine learning model that corresponds to the identified speaker of the translated, extracted speech segment and prosody information for the extracted speech segments; and replace the extracted speech segments from the audio track of the audio/visual file with the spoken versions spoken version of the translated, extracted speech segments to generate a modified audio track.

公开(公告)号：US12272371B1

公开(公告)日：2025-04-08

申请号：US17364805

申请日：2021-06-30

Applicant: Amazon Technologies, Inc.

Inventor： Ritwik Giri ,  Shrikant Venkataramani ,  Jean-Marc Valin ,  Mehmet Umut Isik ,  Arvindh Krishnaswamy

IPC: G06F17/00 ,  G06N20/00 ,  G10L21/013 ,  G10L21/0364 ,  G10L21/038

Abstract: Real-time audio enhancement for a target speaker may be performed. An embedding of a sample of speaker audio is created using a trained neural network that performs voice identification. The embedding is then concatenated with the input features of a trained machine learning model for audio enhancement. The audio enhancement model can recognize and enhance a target speaker's speech in a real-time implementation, as the embedding is in the same feature space of the audio enhancement model.

公开(公告)号：US20250111857A1

公开(公告)日：2025-04-03

申请号：US18478759

申请日：2023-09-29

Applicant: Amazon Technologies, Inc.

Inventor： Ritwik Giri ,  Zhepei Wang ,  Devansh Shah ,  Jean-Marc Valin ,  Michael Mark Goodwin

IPC: G10L21/0208 ,  G10L25/30 ,  H04M3/56

Abstract: Examples herein provide an approach to enhance an audio mixture of a teleconference application by switching between noise suppression modes using a single model. Specifically, a machine learning (ML) model may be configured to, in response to receiving an audio mixture representation as input, suppress either a background noise of the audio mixture or suppress all noise of the audio mixture except a user's voice. In some examples, the ML model may be trained on speech and background noise training data during a training phase. In addition, the ML model may be trained on a user's voice during an enrollment phase. In addition, during an inference phase, the ML model may enhance the audio mixture by suppressing a portion of the audio mixture.

公开(公告)号：US12175434B2

公开(公告)日：2024-12-24

申请号：US17039649

申请日：2020-09-30

Applicant: Amazon Technologies, Inc.

Inventor： Srikanth Venkata Tenneti ,  Arvindh Krishnaswamy ,  Karim Helwani ,  Mehmet Umut Isik ,  Ritwik Giri ,  Fangzhou Cheng ,  Aparna Pandey

IPC: G06Q10/20 ,  G06F16/21 ,  G06F16/906

Abstract: Systems, methods, and apparatuses for detecting anomalies using clusters are described. In some examples, a method includes receiving a request to perform anomaly detection using a plurality of clusters; receiving a data point; determining when the received data point is a part of one of the plurality of clusters utilizing a distance to centers of the one or more clusters, wherein: when the received data point is determined to belong to a normal cluster, assigning the received data point to the determined cluster, updating the cluster, and updating a history for the cluster, when the received data point is determined to belong to an anomalous cluster, raising an anomaly, updating the cluster, and updating a history for the cluster, and when the received data point is determined to not belong to any cluster, raising an anomaly.

公开(公告)号：US12008457B1

公开(公告)日：2024-06-11

申请号：US17037515

申请日：2020-09-29

Applicant: Amazon Technologies, Inc.

Inventor： Mehmet Umut Isik ,  Ritwik Giri ,  Neerad Dilip Phansalkar ,  Jean-Marc Valin ,  Karim Helwani ,  Arvindh Krishnaswamy

IPC: G06N3/045 ,  G06N3/082 ,  G10L15/16

CPC classification number: G06N3/045 ,  G06N3/082 ,  G10L15/16

Abstract: Audio processing may be performed with a convolutional neural network that includes positional embeddings. Audio data may be received at an audio processing system. A convolutional neural network that concatenates frequency-positional embeddings at an input layer may be used to process the audio data. A result of processing the audio data through the convolutional neural network may be used to perform an audio processing task.

公开(公告)号：US20220101270A1

公开(公告)日：2022-03-31

申请号：US17039649

申请日：2020-09-30

Applicant: Amazon Technologies, Inc.

Inventor： Srikanth Venkata Tenneti ,  Arvindh Krishnaswamy ,  Karim Helwani ,  Mehmet Umut Isik ,  Ritwik Giri ,  Fangzhou Cheng ,  Aparna Pandey

IPC: G06Q10/00 ,  G06F16/906 ,  G06F16/21

Abstract: Systems, methods, and apparatuses for detecting anomalies using clusters are described. In some examples, a method includes receiving a request to perform anomaly detection using a plurality of clusters; receiving a data point; determining when the received data point is a part of one of the plurality of clusters utilizing a distance to centers of the one or more clusters, wherein: when the received data point is determined to belong to a normal cluster, assigning the received data point to the determined cluster, updating the cluster, and updating a history for the cluster, when the received data point is determined to belong to an anomalous cluster, raising an anomaly, updating the cluster, and updating a history for the cluster, and when the received data point is determined to not belong to any cluster, raising an anomaly.