公开(公告)号：US20220148293A1

公开(公告)日：2022-05-12

申请号：US17283199

申请日：2020-03-11

Applicant: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY

Inventor： Shuqiang WANG ,  Wen YU ,  Chenchen XIAO ,  Shengye HU ,  Yanyan SHEN

IPC: G06V10/774 ,  G06V10/82 ,  G06N3/04 ,  G06T7/00

Abstract: An image feature visualization method and apparatus, and an electronic device during model training, inputs the real training data with positive samples into a mapping generator to obtain fictitious training data with negative samples. The mapping generator includes a mapping module configured to learn a key feature map that distinguishes the real training data with positive samples/negative samples, and the fictitious training data with negative samples is generated based on the real training data with positive samples and the key feature map. The training data with negative samples is input into a discriminator to obtain a discrimination result. An optimizer optimizes the mapping generator and the discriminator until training is completed. During model application, a target image that is to be processed is input into the mapping generator, and the mapper in the mapping generator extracts features of the target image.

公开(公告)号：US20220343638A1

公开(公告)日：2022-10-27

申请号：US17763513

申请日：2019-11-19

Applicant: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES

Inventor： Shuqiang WANG ,  Wen YU ,  Yanyan SHEN ,  Zhuo CHEN

IPC: G06V10/80 ,  G06V10/82 ,  G06V10/20 ,  G06T7/00 ,  G16H30/40

Abstract: The present application is suitable for use in the technical field of computers, and provides a smart diagnosis assistance method and terminal based on medical images, comprising: acquiring a medical image to be classified; pre-processing the medical image to be classified to obtain a pre-processed image; and inputting the pre-processed image into a trained classification model for classification processing to obtain a classification type corresponding to the pre-processed image, the classification model comprising tensorized network layers and a second-order pooling module. As the trained classification model comprises tensor decomposed network layers and a second-order pooling module, when processing images on the basis of the classification model, more discriminative features related to pathologies can be extracted, increasing the accuracy of medical image classification.

公开(公告)号：US20220101527A1

公开(公告)日：2022-03-31

申请号：US17549258

申请日：2021-12-13

Applicant: Shenzhen Institutes of Advanced Technology

Inventor： Shuqiang WANG ,  Wen YU ,  Chenchen XIAO ,  Shengye HU

IPC: G06T7/00 ,  G06V10/764 ,  G06V10/774

Abstract: A visualization method for evaluating brain addiction traits, an apparatus, and a computer-readable storage medium are provided. The method includes the following. A visualization processing request is received from a client, where the visual processing request contains an image to-be-processed. The image to-be-processed is masked to obtain a perturbation image masked. The perturbation image is classified with a visualization processing model to obtain a classification result, and the classification result is calculated to obtain an evaluation value of the perturbation image, where the evaluation value of the perturbation image is less than an evaluation value of the image to-be-processed without masking. The visualization evaluation result is determined according to the evaluation value of the perturbation image. The visualization evaluation result is sent to the client.