公开(公告)号：US12126171B1

公开(公告)日：2024-10-22

申请号：US17116487

申请日：2020-12-09

Applicant: Amazon Technologies, Inc.

Inventor： Nayana Teja Chiluvuri ,  Chang Hwa Rob Yang ,  Xunwei Yu ,  Kenneth Lawrence Staton ,  Cameron Dean Whitehouse ,  Matthew L. Chaboud ,  Sajesh Kumar Saran ,  Pushkaraksha Gejji ,  Krzysztof Marcin Walczak

IPC: H02J13/00 ,  G06Q50/06 ,  H02J3/14

CPC classification number: H02J3/14 ,  G06Q50/06 ,  H02J13/00017

Abstract: Described is an apparatus to concurrently measure multiple input voltages at a high sampling data rate, such as at least two mega-samples per second. The apparatus may include a plurality of voltage data acquisition components that concurrently sample different input voltages and produce respective voltage data samples. Each of the plurality of voltage data acquisition components may be directly coupled to a field programmable gate array that receives the voltage data samples, packetizes those voltage data samples, and provide the packetized voltage data samples to a system on chip.

公开(公告)号：US11321580B1

公开(公告)日：2022-05-03

申请号：US16673769

申请日：2019-11-04

Applicant: Amazon Technologies, Inc.

Inventor： Krzysztof Marcin Walczak ,  Emilio Ian Maldonado ,  Bella Dubrov

IPC: G06K9/62 ,  G06F16/957 ,  G06F40/284 ,  G06N3/08 ,  G06N3/04

Abstract: Systems and methods are provided for learning item types of items listed in an electronic repository, and for training a machine learning model to predict the item type of a given input item. For example, a machine learning model may be obtained or accessed that has been previously trained to classify an input item to a browse node. Vector representations of individual items assigned to different browse nodes may be obtained from an intermediate layer of the previously trained machine learning model, and a vector representation of individual browse nodes may then be generated based on the vector representations of individual items assigned to that browse node. A clustering algorithm may be applied to the browse node vector representations in order to identify clusters of similar browse nodes, where individual clusters may represent different unique item types.