-
公开(公告)号：US11526693B1
公开(公告)日：2022-12-13
申请号：US16865167
申请日：2020-05-01
Applicant: Amazon Technologies, Inc.
Inventor： Roman Goldenberg , Miriam Farber , George Leifman , Gerard Guy Medioni
Abstract: Disclosed are systems and method for training an ensemble of machine learning models with a focus on feature engineering. For example, the training of the models encourages each machine learning model of the ensemble to rely on a different set of input features from the training data samples used to train the machine learning models of the ensemble. However, instead of telling each model explicitly which features to learn, in accordance with the disclosed implementations, ML models of the ensemble may be trained sequentially, with each new model trained to disregard input features learned by previously trained ML models of the ensemble and learn based on other features included in the training data samples.
-
Patent Agency Ranking
公开(公告)号：US11900711B1
公开(公告)日：2024-02-13
申请号：US17075926
申请日：2020-10-21
Applicant: AMAZON TECHNOLOGIES, INC.
Inventor： Dilip Kumar , Manoj Aggarwal , George Leifman , Gerard Guy Medioni , Nikolai Orlov , Natan Peterfreund , Korwin Jon Smith , Dmitri Veikherman , Sora Kim
CPC classification number: G06V40/1318 , G06N3/045 , G06V40/1347 , G06T2207/20084 , G06V40/1341
Abstract: An identification system includes one or more infrared light sources and a camera that acquires images of a user's palm. For example, at a first time, one or more first images may be acquired by the camera using infrared light with a first polarization that represent external characteristics of the user's palm. At a second time, one or more second images may be acquired using infrared light with a second polarization that represent internal characteristics of the user's palm. These images are processed to determine a first set of feature vectors and a second set of feature vectors. A current signature may be determined using the first set of feature vectors and the second set of feature vectors. In addition, a user may be identified based on a comparison of the current signature and previously stored reference signatures that are associated with candidate user identifiers.
-
公开(公告)号：US11328513B1
公开(公告)日：2022-05-10
申请号：US15806098
申请日：2017-11-07
Applicant: Amazon Technologies, Inc.
Inventor： Eli Osherovich , Ehud Benyamin Rivlin , Yacov Hel-Or , Dmitri Veikherman , Dilip Kumar , Gerard Guy Medioni , George Leifman
Abstract: Described is a multiple-camera system and process for detecting, tracking, and re-verifying agents within a materials handling facility. In one implementation, a plurality of feature vectors may be generated for an agent and maintained as an agent model representative of the agent. When the object being tracked as the agent is to be re-verified, feature vectors representative of the object are generated and stored as a probe agent model. Feature vectors of the probe agent model are compared with corresponding feature vectors of candidate agent models for agents located in the materials handling facility. Based on the similarity scores, the agent may be re-verified, it may be determined that identifiers used for objects tracked as representative of the agents have been flipped, and/or to determine that tracking of the object representing the agent has been dropped.
-
公开(公告)号：US20190392189A1
公开(公告)日：2019-12-26
申请号：US16014843
申请日：2018-06-21
Applicant: AMAZON TECHNOLOGIES, INC
Inventor： Dilip Kumar , Manoj Aggarwal , George Leifman , Gerard Guy Medioni , Nikolai Orlov , Natan Peterfreund , Korwin Jon Smith , Dmitri Veikherman , Sora Kim
Abstract: A non-contact biometric identification system includes a hand scanner that generates images of a user's palm. Images are acquired using light of a first polarization at a first time show surface characteristics such as wrinkles in the palm while images acquired using light of a second polarization at a second time show deeper characteristics such as veins. Within the images, the palm is identified and subdivided into sub-images. The sub-images are subsequently processed to determine feature vectors present in each sub-image. A current signature is determined using the feature vectors. A user may be identified based on a comparison of the current signature with a previously stored reference signature that is associated with a user identifier.
-
公开(公告)号：US11961303B1
公开(公告)日：2024-04-16
申请号：US17738960
申请日：2022-05-06
Applicant: Amazon Technologies, Inc.
Inventor： Eli Osherovich , Ehud Benyamin Rivlin , Yacov Hel-Or , Dmitri Veikherman , Dilip Kumar , Gerard Guy Medioni , George Leifman
CPC classification number: G06V20/52 , G06F18/22 , G06V10/74 , G06V10/751 , G06V40/103 , G06V40/166 , G06V40/168 , G06V40/172 , G06V40/23
Abstract: Described is a multiple-camera system and process for detecting, tracking, and re-verifying agents within a materials handling facility. In one implementation, a plurality of feature vectors may be generated for an agent and maintained as an agent model representative of the agent. When the object being tracked as the agent is to be re-verified, feature vectors representative of the object are generated and stored as a probe agent model. Feature vectors of the probe agent model are compared with corresponding feature vectors of candidate agent models for agents located in the materials handling facility. Based on the similarity scores, the agent may be re-verified, it may be determined that identifiers used for objects tracked as representative of the agents have been flipped, and/or to determine that tracking of the object representing the agent has been dropped.
-
公开(公告)号：US11636286B1
公开(公告)日：2023-04-25
申请号：US16865187
申请日：2020-05-01
Applicant: Amazon Technologies, Inc.
Inventor： Roman Goldenberg , Miriam Farber , George Leifman , Gerard Guy Medioni
Abstract: Described are systems and methods for training machine learning models of an ensemble of models that are de-correlated. For example, two or more machine learning models may be concurrently trained (e.g., co-trained) while adding a decorrelation component to one or both models that decreases the pairwise correlation between the outputs of the models. Unlike traditional approaches, in accordance with the disclosed implementations, only the negative results need to be decorrelated.
-
-
-
-
-
Search Results
6
records
(took 0.02 seconds)
