公开(公告)号：US10360482B1

公开(公告)日：2019-07-23

申请号：US15830952

申请日：2017-12-04

Applicant: Amazon Technologies, Inc.

Inventor： Vineet Khare ,  Gurumurthy Swaminathan ,  Xiong Zhou

IPC: G06K9/62 ,  G06K9/66 ,  G06N3/04 ,  G06N3/08

Abstract: Features related to systems and methods for generating a machine learning model that is a composite of at least two other models (e.g., crowd-sourced models contributed by users) are described. Each of the contributed models provide output values that may not be to scale. To account for these differences, a normalization factor for a first machine learning model is generated to adjust values produced by the first machine learning model to correspond with results from the second machine learning model. The crowd-sourced models along with the normalization factor are included in the new image model generated in the claims.

公开(公告)号：US12277192B1

公开(公告)日：2025-04-15

申请号：US16560814

申请日：2019-09-04

Applicant: Amazon Technologies, Inc.

Inventor： Ragav Venkatesan ,  Xiong Zhou ,  Gurumurthy Swaminathan ,  Fedor Zhdanov

IPC: G06F18/214 ,  G06N3/08 ,  G06N5/04 ,  G06N20/00

Abstract: Techniques for zero-shot and few-shot transfer of domain-adapted base networks are described. Multiple machine learning task layers are trained using a shared base feature extractor network. At least one task layer is trained with samples and corresponding labels from a first domain as well as a second domain. At least one other task layer is trained with samples and corresponding labels from only the first domain. Ultimately, the other task layer(s) are adapted to generate labels for the first domain via the base network being weighted based on all trainings.

公开(公告)号：US10474926B1

公开(公告)日：2019-11-12

申请号：US15815492

申请日：2017-11-16

Applicant: Amazon Technologies, Inc.

Inventor： Leo Parker Dirac ,  Vineet Khare ,  Gurumurthy Swaminathan ,  Xiong Zhou

IPC: G06K9/62 ,  G06N3/04 ,  G06N3/08

Abstract: Features related to systems and methods expediting generation of a machine learning model, such as an image recognition model, are described. Existing machine learning models are analyzed to identify a starting point for creating the new machine learning model. An existing machine learning model can suggest learning parameters (e.g., training parameters or structural features of the model) that can be used to expedite the generating and training process along with training data that can augment the training of the new machine learning model.

公开(公告)号：US11809992B1

公开(公告)日：2023-11-07

申请号：US16836376

申请日：2020-03-31

Applicant: Amazon Technologies, Inc.

Inventor： Gurumurthy Swaminathan ,  Ragav Venkatesan ,  Xiong Zhou ,  Runfei Luo ,  Vineet Khare

IPC: G06N3/08 ,  G06N3/082 ,  G06N3/045

CPC classification number: G06N3/082 ,  G06N3/045

Abstract: Neural networks with similar architectures may be compressed using shared compression profiles. A request to compress a trained neural network may be received and an architecture of the neural network identified. The identified architecture may be compared with the different network architectures mapped to compression profiles to select a compression profile for the neural network. The compression profile may be applied to remove features of the neural network to generate a compressed version of the neural network.

公开(公告)号：US11501173B1

公开(公告)日：2022-11-15

申请号：US16831595

申请日：2020-03-26

Applicant: Amazon Technologies, Inc.

Inventor： Gurumurthy Swaminathan ,  Ragav Venkatesan ,  Xiong Zhou ,  Runfei Luo ,  Vineet Khare

IPC: G06F15/00 ,  G06F3/12 ,  G06K1/00 ,  G06N5/00 ,  G06N20/00

Abstract: A compression policy to produce compression profiles for compressing trained machine learning models may be trained using reinforcement learning. An iterative reinforcement learning may be performed response to a search request. Different prospective compression profiles may be generated for received machine learning models according to a compression policy being trained. Performance of compressed versions of the trained neural networks according to the compression profiles may be caused using data sets used to train the machine learning models. The compression policy may be updated according to reward signal determined from an application of a reward function for performance criteria to performance results of the different versions of the machine learning models. When a search criteria is satisfied, the trained compression policy may be provided.

公开(公告)号：US20230409584A1

公开(公告)日：2023-12-21

申请号：US18334091

申请日：2023-06-13

Applicant: Amazon Technologies, Inc.

Inventor： Ragav Venkatesan ,  Gurumurthy Swaminathan ,  Xiong Zhou ,  Runfei Luo ,  Vineet Khare

IPC: G06F16/2458 ,  G06F16/25 ,  G06F16/248 ,  G06N3/04 ,  H03M7/30 ,  G06N3/082

CPC classification number: G06F16/2474 ,  G06F16/252 ,  G06F16/248 ,  G06N3/04 ,  H03M7/30 ,  G06N3/082 ,  G06Q10/10

Abstract: Compression profiles may be searched for trained neural networks. An iterative compression profile search may be performed response to a search request. Different prospective compression profiles may be generated for trained neural networks according to a search policy. Performance of compressed versions of the trained neural networks according to the compression profiles may be tracked. The search policy may be updated according to an evaluation of the performance of the compression profiles for the compressed versions of the trained neural networks using compression performance criteria. When a search criteria is satisfied, a result for the compression profile search may be provided.

公开(公告)号：US11755603B1

公开(公告)日：2023-09-12

申请号：US16831584

申请日：2020-03-26

Applicant: Amazon Technologies, Inc.

Inventor： Ragav Venkatesan ,  Gurumurthy Swaminathan ,  Xiong Zhou ,  Runfei Luo ,  Vineet Khare

IPC: G06F16/24 ,  G06F16/2458 ,  G06F16/25 ,  G06F16/248 ,  G06N3/04 ,  H03M7/30 ,  G06N3/082 ,  G06Q10/10

CPC classification number: G06F16/2474 ,  G06F16/248 ,  G06F16/252 ,  G06N3/04 ,  G06N3/082 ,  H03M7/30 ,  G06Q10/10

Abstract: Compression profiles may be searched for trained neural networks. An iterative compression profile search may be performed response to a search request. Different prospective compression profiles may be generated for trained neural networks according to a search policy. Performance of compressed versions of the trained neural networks according to the compression profiles may be tracked. The search policy may be updated according to an evaluation of the performance of the compression profiles for the compressed versions of the trained neural networks using compression performance criteria. When a search criteria is satisfied, a result for the compression profile search may be provided.