Huawei Technologies Co., Limited

Shenzhen Institutes of Advanced Technology

Foxconn Technology Co., Ltd.

Zte Corporation

Fu Zhun Precision Industry (shen Zhen) Co., Ltd.

Na Zhang

Jianhua Zheng

Jicheng An

Xin Liu

Dong Liang

Zhanli Hu

Tay-Jian Liu

Ping Li

Hairong Zheng

Yongfeng Yang

Quanhe Yu

Bin Gao

Qi Yang

Jian-Bing Qian

Qihang Chen

Hanyu Sun

Zhiyuan Huang

A medical image noise reduction method, system, terminal and storage medium are disclosed. The method includes: obtaining a standard-dose PET image and a constant-value image; inputting the standard-dose PET image and the constant-value image into a decay function to obtain a corresponding low-dose noisy PET image and a noisy constant-value image, respectively; assembling the low-dose noisy PET image and the noisy constant-value image in a width dimension or a height dimension, and then inputting into a trained conjugate generative adversarial network, and outputting a denoised PET image and constant-value image output by the conjugate generative adversarial network.

Medical image noise reduction method, system, terminal, and storage medium

A medical image processing method and processing apparatus, and a computer readable storage medium. The method includes: obtaining a to-be-processed image; performing a feature extraction on the to-be-processed image to obtain a corresponding feature image; and re-determining a pixel value of each pixel in the to-be-processed image based on first information and second information of a corresponding pixel in the feature image, and processing the to-be-processed image; wherein the first information is information of a pixel adjacent to the corresponding pixel in the features image, and the second information is information of a pixel that is not adjacent to and is similar to the corresponding pixel in the features image.

Medical image processing method, processing apparatus, and computer-readable storage medium

Methods and apparatuses for training a magnetic resonance imaging model, electronic devices and computer readable storage media are provided. A method may include: acquiring a magnetic resonance image data set; constructing a ring deep neural network to be trained; inputting an under-sampled magnetic resonance image and a full-sampled magnetic resonance image respectively to two neural networks included in the ring deep neural network, to generate respective simulated magnetic resonance images; inputting a first simulated full-sampled magnetic resonance image and the full-sampled magnetic resonance image to a pre-constructed first simulated magnetic resonance image class discrimination model, to obtain a first discrimination result indicating whether or not the first simulated full-sampled magnetic resonance image is of a simulated magnetic resonance image class; and adjusting a network parameter of the ring deep neural network based on a preset loss function, to obtain a trained magnetic resonance imaging model.

Methods and apparatuses for training magnetic resonance imaging model

An inter prediction method is provided, including: obtaining a first reference frame index and a first motion vector of a to-be-processed block (); deriving a second reference frame index from the first reference frame index, where the second reference frame index is different from the first reference frame index (); scaling the first motion vector to obtain a second motion vector (); obtaining a first reference block of the to-be-processed block according to the first reference frame index and the first motion vector (); obtaining a second reference block of the to-be-processed block according to the second reference frame index and the second motion vector (); and weighting the first reference block and the second reference block to obtain a combined prediction block (). Each reference list for inter prediction is allowed to include two groups of motion information, which helps improve an inter prediction effect.

Inter prediction method and apparatus

A method for predicting motion information of a picture block and a related product is provided. The method includes: determining a candidate motion information list of a current picture block, where the candidate motion information list includes at least one combined candidate motion information, one combined candidate motion information is obtained by weighting P candidate motion information by using corresponding weighting factors; determining target motion information in the candidate motion information list; and predicting motion information of the current picture block based on the target motion information.

Method and apparatus for predicting motion information of picture block, encoder, and decoder

The present application relates to a method and system for generating multi-task learning-type generative adversarial network for low-dose PET reconstruction, and relates to the field of deep learning. The method includes connecting layers of the encoder with layers of the decoder by skip connection to provide a U-Net type picture generator; generating a group of generative adversarial networks by matching a plurality of picture generators with a plurality of discriminators in one-to-one manner; obtaining a first multi-task learning-type generative adversarial network; designing a joint loss functionfor improving image quality; and training the first multi-task learning-type generative adversarial network according to the joint loss functionin combination with an optimizer to provide a second multi-task learning-type generative adversarial network.

Method and system for generating multi-task learning-type generative adversarial network for low-dose PET reconstruction

A method and an apparatus and for intra prediction of a picture block are provided, to provide a manner of predicting a current block by using spatial non-adjacent blocks or a temporal neighboring block, thereby improving coding performance. In various embodiments, a first luma intra mode set is constructed; and a bitstream is parsed to obtain a first identifier. In those embodiments, when the first identifier indicates that a luma prediction mode of the current block belongs to the first luma intra mode set, the bitstream is parsed to obtain a second identifier. The second identifier indicates that a candidate luma prediction mode in the first luma intra mode set is used as the luma prediction mode of the current block. Still in those embodiments, a luma prediction value of the current block is obtained based on the luma prediction mode of the current block.

Method and apparatus for intra prediction of picture block

Inter prediction method and apparatus for video coding

A video picture inter prediction method includes determining an inter prediction mode used to perform inter prediction on a current picture block, where the inter prediction mode is one mode in a candidate inter prediction mode set, and the candidate inter prediction mode set includes a plurality of inter prediction modes used for a non-directional motion field or a plurality of inter prediction modes used for a directional motion field, and performing inter prediction on the current picture block based on the determined inter prediction mode.

Video picture inter prediction method and apparatus, and codec

A method of processing plaques in magnetic resonance imaging of vessel wall include: step Straining a generative adversarial network and a capsule neural network to obtain a trained generator network and a trained capsule neural network; and step Scascade-connecting the trained generator network with the capsule neural network into a system to recognize and classify plaques in magnetic resonance imaging of vessel wall. In one aspect, the capsule neural network has more abundant vascular plaques characteristic information represented by vector; in another aspect, when the trained generator network and the capsule neural network are cascaded into the system to recognize and classify the plaques in magnetic resonance imaging of vessel wall, an accuracy of recognition and classification may be greatly improved. A device for processing the method as well as a computer for implementing are also disclosed.

Method and device of processing plaques in magnetic resonance imaging of vessel wall, and computer device

The disclosure discloses a method and device for sending and receiving data, which are used to solve the problem that the existing device having the function of the laser pen and the function of the data card cannot implement the integrated control over the function of the laser pen and the function of the data card. The device includes a main module and a connection module. The connection module is configured to connect to a PC by using a USB mode, and the main module communicates with the connection module by high frequency. The method for sending data includes: the main module encapsulates data to be sent according to a type of the data to be sent and identification information corresponding to the type of the data, and sends the encapsulated data to the connection module; and the connection module decapsulates the encapsulated data and sends the decapsulated data to the PC. Since the device determines the identification information corresponding to the type according to the type of the received data, and encapsulates the identification information and the data, in order to be capable of performing corresponding processing when needed, so that intelligent integration of the function of the laser pen and the function of the data card is implemented and an intelligent control over the function of the laser pen is implemented.

Method and device for sending and receiving data

A method and system for configuring a complex network and a server resource configuration and management module are provided. The method comprises: obtaining an access record of each cloud computing server in a cloud server pool in a network; constructing a complex network model according to the access record(s); performing a community division process to the complex network model to obtain a community structure and memberships of members in each community; and configuring a structure of the network according to the community structure and the memberships of the members in each community. The provided technical scheme is applicable to a cloud computing network. The method and system for configuring a complex network and the server resource configuration and management module, realize configuring the network according to the usage of each cloud computing server in the current network, solve the problem of no mechanism for configuring a cloud computing network reasonably.

Configuration method and system of complex network and configuration and management module of server resources

An LED illumination device includes a heat dissipation part, an optical part and an electric part. The optical part includes an LED module attached on the heat dissipation part and an envelope mounted on the heat dissipation part and covering the LED module. The electric part includes a casing and a lamp cap. The casing is cup-shaped. The lamp cap includes a sleeve electrode and a spring electrode attached at a bottom of the sleeve electrode. The sleeve electrode and the spring electrode electrically power the LED module. One end of the casing along an axis thereof is mounted on bottom ends of the heat dissipation part and the optical part. The sleeve electrode is disposed around the other end of the casing. The spring electrode is resiliently deformable along an axis of the sleeve electrode.

LED illumination device

An LED illuminating device includes an optical part having a light source therein, an electrical part and a heat dissipation part. The electrical part includes a rear shell and a circuit board electrically connecting with the light source. The heat dissipation part includes a hollow front shell having a rear end secured with the rear shell, a heat sink and a cooling fan arranged in the front shell. The front shell defines a plurality of first and second air holes. The heat sink includes a solid base and a plurality of fins. The light source is thermally connected with the base. An airflow produced by the cooling fan is introduced into the front shell via the first air holes. The airflow flows then towards the heat sink and finally flows out of the front shell via the second air holes.

Growing community of inventors

Shenzhen, China

Na Zhang