公开(公告)号：US20220058437A1

公开(公告)日：2022-02-24

申请号：US16999665

申请日：2020-08-21

Applicant: GE Precision Healthcare LLC

Inventor： Ravi Soni ,  Tao Tan ,  Gopal B. Avinash ,  Dibyaiyoti Pati ,  Hans Krupakar ,  Venkata Ratnam Saripalli

IPC: G06K9/62 ,  G06N20/00

Abstract: Systems and techniques that facilitate synthetic training data generation for improved machine learning generalizability are provided. In various embodiments, an element augmentation component can generate a set of preliminary annotated training images based on an annotated source image. In various aspects, a preliminary annotated training image can be formed by inserting at least one element of interest or at least one background element into the annotated source image. In various instances, a modality augmentation component can generate a set of intermediate annotated training images based on the set of preliminary annotated training images. In various cases, an intermediate annotated training image can be formed by varying at least one modality-based characteristic of a preliminary annotated training image. In various aspects, a geometry augmentation component can generate a set of deployable annotated training images based on the set of intermediate annotated training images. In various instances, a deployable annotated training image can be formed by varying at least one geometric characteristic of an intermediate annotated training image. In various embodiments, a training component can train a machine learning model on the set of deployable annotated training images.

公开(公告)号：US11720647B2

公开(公告)日：2023-08-08

申请号：US16999665

申请日：2020-08-21

Applicant: GE Precision Healthcare LLC

Inventor： Ravi Soni ,  Tao Tan ,  Gopal B. Avinash ,  Dibyajyoti Pati ,  Hans Krupakar ,  Venkata Ratnam Saripalli

IPC: G06F18/214 ,  G06N20/00 ,  G06F18/21 ,  G06N3/08 ,  G06N20/10 ,  G06V10/774 ,  G16H30/40 ,  G06T11/00

CPC classification number: G06F18/2148 ,  G06F18/214 ,  G06F18/2163 ,  G06N3/08 ,  G06N20/00 ,  G06N20/10 ,  G06T11/00 ,  G06V10/774 ,  G16H30/40 ,  G06V2201/03

Abstract: Systems and techniques that facilitate synthetic training data generation for improved machine learning generalizability are provided. In various embodiments, an element augmentation component can generate a set of preliminary annotated training images based on an annotated source image. In various aspects, a preliminary annotated training image can be formed by inserting at least one element of interest or at least one background element into the annotated source image. In various instances, a modality augmentation component can generate a set of intermediate annotated training images based on the set of preliminary annotated training images. In various cases, an intermediate annotated training image can be formed by varying at least one modality-based characteristic of a preliminary annotated training image. In various aspects, a geometry augmentation component can generate a set of deployable annotated training images based on the set of intermediate annotated training images. In various instances, a deployable annotated training image can be formed by varying at least one geometric characteristic of an intermediate annotated training image. In various embodiments, a training component can train a machine learning model on the set of deployable annotated training images.

公开(公告)号：US20220253708A1

公开(公告)日：2022-08-11

申请号：US17174049

申请日：2021-02-11

Applicant: GE Precision Healthcare LLC

Inventor： Rajesh Kumar Tamada ,  Junpyo Hong ,  Attila Márk Rádics ,  Hans Krupakar ,  Venkata Ratnam Saripalli ,  Dibyajyoti Pati ,  Guarav Kumar

IPC: G06N3/08 ,  G06N3/063 ,  G06N5/04 ,  G06K9/62

Abstract: Techniques are provided for compressing deep neural networks using a structured filter pruning method that is extensible and effective. According to an embodiment, a computer-implemented method comprises determining, by a system operatively coupled to a processor, importance scores for filters of layers of a neural network model previously trained until convergence for an inferencing task on a training dataset. The method further comprises removing, by the system, a subset of the filters from one or more layers of the layers based on the importance scores associated with the subset failing to satisfy a threshold importance score value. The method further comprises converting, by the system, the neural network model into a compressed neural network model with the subset of the filters removed.