-
1.发明授权公开(公告)号：US11257002B2
公开(公告)日：2022-02-22
申请号：US15919628
申请日：2018-03-13
Applicant: Amazon Technologies, Inc.
Inventor： Thomas Albert Faulhaber, Jr. , Edo Liberty , Stefano Stefani , Zohar Karnin , Craig Wiley , Steven Andrew Loeppky , Swaminathan Sivasubramanian , Alexander Johannes Smola , Taylor Goodhart
Abstract: Techniques for dynamic accuracy-based experimentation and deployment of machine learning (ML) models are described. Inference traffic flowing to ML models and the accuracy of the models is analyzed and used to ensure that better performing models are executed more often via model selection. A predictive component can evaluate which model is more likely to be accurate for certain input data elements. Ensemble techniques can combine inference results of multiple ML models to aim to achieve a better overall result than any individual model could on its own.
-
公开(公告)号：US20180174595A1
公开(公告)日：2018-06-21
申请号：US15387038
申请日：2016-12-21
Applicant: Amazon Technologies, Inc.
Inventor： Leo Parker Dirac , Fabian Moerchen , Edo Liberty
IPC: G10L21/013 , G10L15/00 , G10L15/16 , G10L25/51
CPC classification number: G10L21/013 , G06F17/289 , G10L15/00 , G10L15/005 , G10L25/51
Abstract: Techniques for accent translation are described herein. A plurality of audio samples may be received, and each of the plurality of audio samples may be associated with at least one of a plurality of accents. Audio samples associated with at least a first accent of the plurality of accents may be compared to audio samples associated with at least one other accent of the plurality of accents. A translation model between the first accent and a second accent may be generated. An input audio portion in a first spoken language may be received. It may be determined whether the input audio portion is substantially associated with the first accent, and if so, an output audio portion substantially associated with the second accent in the first spoken language may be outputted based, at least in part, on the translation model.
-
公开(公告)号：US11176489B1
公开(公告)日：2021-11-16
申请号：US15954913
申请日：2018-04-17
Applicant: Amazon Technologies, Inc.
Inventor： Alexander Johannes Smola , Edo Liberty , Mu Li , Leyuan Wang
Abstract: Techniques for determining and utilizing optimal aggregation schedules are described are described. A deep machine learning model can be trained using multiple processing elements implemented in one or multiple computing devices and that are interconnected using one or multiple types of links. An optimal aggregation schedule for such arbitrary topologies can be determined automatically. The determination may include solving a linear program on the spanning tree polytope. The optimal aggregation schedule can be utilized by the multiple processing elements to train the deep machine learning model.
-
公开(公告)号：US10163451B2
公开(公告)日：2018-12-25
申请号：US15387038
申请日：2016-12-21
Applicant: Amazon Technologies, Inc.
Inventor： Leo Parker Dirac , Fabian Moerchen , Edo Liberty
IPC: G10L21/013 , G10L15/00 , G10L25/51 , G06F17/28
Abstract: Techniques for accent translation are described herein. A plurality of audio samples may be received, and each of the plurality of audio samples may be associated with at least one of a plurality of accents. Audio samples associated with at least a first accent of the plurality of accents may be compared to audio samples associated with at least one other accent of the plurality of accents. A translation model between the first accent and a second accent may be generated. An input audio portion in a first spoken language may be received. It may be determined whether the input audio portion is substantially associated with the first accent, and if so, an output audio portion substantially associated with the second accent in the first spoken language may be outputted based, at least in part, on the translation model.
-
公开(公告)号：US12277480B1
公开(公告)日：2025-04-15
申请号：US15934091
申请日：2018-03-23
Applicant: Amazon Technologies, Inc.
Inventor： Edo Liberty , Thomas Albert Faulhaber, Jr. , Zohar Karnin , Gowda Dayananda Anjaneyapura Range , Amir Sadoughi , Swaminathan Sivasubramanian , Alexander Johannes Smola , Stefano Stefani , Craig Wiley
Abstract: Techniques for in-flight scaling of machine learning training jobs are described. A request to execute a machine learning (ML) training job is received within a provider network, and the ML training job is executed using a first one or more compute instances. Upon a determination that a performance characteristic of the ML training job satisfies a scaling condition, a second one or more compute instances are added to the ML training job while the first one or more compute instances continue to execute portions of the ML training job.
-
Patent Agency Ranking
公开(公告)号：US11126927B2
公开(公告)日：2021-09-21
申请号：US15822061
申请日：2017-11-24
Applicant: Amazon Technologies, Inc.
Inventor： Stefano Stefani , Steven Andrew Loeppky , Thomas Albert Faulhaber, Jr. , Craig Wiley , Edo Liberty
Abstract: Techniques for auto-scaling hosted machine learning models for production inference are described. A machine learning model can be deployed in a hosted environment such that the infrastructure supporting the machine learning model scales dynamically with demand so that performance is not impacted. The model can be auto-scaled using reactive techniques or predictive techniques.
-
7.发明授权公开(公告)号：US10824913B1
公开(公告)日：2020-11-03
申请号：US16198313
申请日：2018-11-21
Applicant: Amazon Technologies, Inc.
Inventor： Sahika Genc , Edo Liberty
Abstract: Techniques for performing image-augmentation based simulations on are described. An exemplary embodiment of such performances includes for each tuple of timestamped image and movement data, generating a next image using an image generation neural network based on the timestamped image and movement data, the image being input into the image generation neural network as a non-rendered image, and generating a reward using a reward generating neural network based on the timestamped image and movement data.
-
公开(公告)号：US12045693B2
公开(公告)日：2024-07-23
申请号：US16001548
申请日：2018-06-06
Applicant: Amazon Technologies, Inc.
Inventor： Charles Drummond Swan , Edo Liberty , Steven Andrew Loeppky , Stefano Stefani , Alexander Johannes Smola , Swaminathan Sivasubramanian , Craig Wiley , Richard Shawn Bice , Thomas Albert Faulhaber, Jr. , Taylor Goodhart
CPC classification number: G06N20/00 , G06F9/45558 , G06F2009/45595
Abstract: Techniques for using scoring algorithms utilizing containers for flexible machine learning inference are described. In some embodiments, a request to host a machine learning (ML) model within a service provider network on behalf of a user is received, the request identifying an endpoint to perform scoring using the ML model. An endpoint is initialized as a container running on a virtual machine based on a container image and used to score data and return a result of said scoring to a user device.
-
公开(公告)号：US11537439B1
公开(公告)日：2022-12-27
申请号：US15934046
申请日：2018-03-23
Applicant: Amazon Technologies, Inc.
Inventor： Edo Liberty , Thomas Albert Faulhaber, Jr. , Zohar Karnin , Gowda Dayananda Anjaneyapura Range , Amir Sadoughi , Swaminathan Sivasubramanian , Alexander Johannes Smola , Stefano Stefani , Craig Wiley
Abstract: Techniques for intelligent compute resource selection and utilization for machine learning training jobs are described. At least a portion of a machine learning (ML) training job is executed a plurality of times using a plurality of different resource configurations, where each of the plurality of resource configurations includes at least a different type or amount of compute instances. A performance metric is measured for each of the plurality of the executions, and can be used along with a desired performance characteristic to generate a recommended resource configuration for the ML training job. The ML training job is executed using the recommended resource configuration.
-
公开(公告)号：US10482887B1
公开(公告)日：2019-11-19
申请号：US15925451
申请日：2018-03-19
Applicant: Amazon Technologies, Inc.
Inventor： Madhav Jha , Edo Liberty
Abstract: Techniques for using machine learning models to approximate a user in a communication are described. For example, a method of initiating a communication link with an edge device to exchange audio data; receiving a compressed audio data from the edge device; re-encoding the received compressed audio data using a re-encoding machine learning model to approximate a voice; and outputting the approximated voice is detailed.
-
-
-
-
-
-
-
-
-
Search Results
10
records
(took 0.016 seconds)
