Hong Kong Applied Science and Technology Research Institute Company, Limited

Shenzhen University

Guangdong Oppo Mobile Telecommunications Corp., Ltd.

Byd Company Limited

Tencent Technology (shenzhen) Company Limited

Shenzhen Mindray Bio-medical Electronics Co., Ltd.

The Hong Kong University of Science and Technology

Hong Kong Applied Science and Technology Research

Institute Company Limited

Shenzhen Unnersity

China Mobile Information Technology Co., Ltd

Huawei Technologies Co., Limited

Lu Wang

Kaishun Wu

Daolin Li

Dong Ma

Ming Li

Hui Yuan

Yongpan Zou

Xiaohui Wang

Xuejian He

Xiao Yuan

Zuyue Xu

Rong Li

Jun Xu

Jian Yu

Li Zhang

Jing Liu

Hong Liu

Hui Zhang

Yan Huo

Ting Wang

Chi Keung Fong

Huchuan Lu

Mulong Liu

King Ngi Ngan

Zhehui Wu

Xiaohua Wu

Min Sun

He Wang

Jun Ge

Qi Luo

Ping Shun Leung

Shangping Liu

Xingde Chi

Hao Lu

Xinwan Wu

Hoi Ming Wong

Jie Dong

Wai-Kuen Cham

Dan Wang

Kaixin Chen

Jinyong Wu

Yaping Qiu

Zhiwei Su

Zhuojun Hua

Yongzhi Huang

Qiang Wan

Xiaoshen Li

Fengfeng Liang

Ziyu Ni

Jianhao Zhang

Ruifeng Huang

Thomas H Y Pun

Tao Xia

Pingping Zhang

Wei Hu

Yujun Li

Lingfeng Xu

Oscar C Au

Sun Wenxiu

Fei Yu

Vinit Jakhetiya

Hailiang Yang

Yu Hu

Su Yang

Jiaqi Zhuang

Qianwen Cao

Congqiang Xu

The present application discloses a deep learning method and system for spectrum sharing of partially overlapping channels. The method comprises the following steps: in response to a received user transmission request, inputting, by a base station, channel state information CSI of a plurality of historical time slices into a trained channel prediction convolutional neural network model and outputting predicted channel state information CSI of a next time slice; and inputting the channel state information CSI of the next time slice into a reinforcement learning model and obtaining a channel allocation strategy of each user equipment in a collision domain of the base station so as to realize a maximum throughput of simultaneous communication of each user equipment, wherein the reinforcement learning model is obtained by training by taking bandwidth efficiency performance as a reward. The present application is high in universality, bandwidth utilization rate and throughput for communication networks.

Deep learning method and system for spectrum sharing of partially overlapping channels

A point cloud encoding method, including: obtaining attribute information about the attributes of a current point in a point cloud; processing the attribute information of the current point to obtain a residual value of the attribute information of the current point; using a target quantization manner, quantizing the residual value of the attribute information of the current point to obtain a quantized residual value of the attribute information of the current point; the target quantization manner comprises at least two of the following quantization manners: a first quantization manner, a second quantization manner, and a third quantization manner.

Point cloud encoding and decoding method and system, and point cloud encoder and point cloud decoder

The invention provides a low-power wide area network integrated sensing and communication method based on channel sensing and reinforcement learning, and the method comprises the steps: receiving an uplink signal transmitted by a terminal node through a gateway receiver, demodulating the uplink signal through employing a soft demodulation algorithm, and obtaining a bit log-likelihood ratio; the deep reinforcement learning model completes channel reconstruction of the corresponding link during communication according to the bit log-likelihood ratio; and calculating the current optimal network configuration according to the reconstructed channel so as to distribute the current optimal network configuration to each terminal node during next downlink transmission. According to the method, the key environment information is mined by using the deep reinforcement learning model, so that the gateway can allocate optimal network configuration to different terminal nodes according to the current network environment condition, and the decoding capability of the gateway is improved.

Low-power wide area network integrated sensing and communication method based on channel sensing and reinforcement learning

A point cloud coding method, including: partitioning a point cloud, and determining a current coding block; determining a quantization parameter optimization enable identifier, a hierarchical level index, and a quantization parameter offset parameter of the current coding block; determining, according to the hierarchical level index and the quantization parameter offset parameter, a quantization step Qstep corresponding to a level of detail; and upon detecting that the quantization parameter optimization enable identifier is a first value, performing operations for one or more points included in the current coding block.

Point cloud coding method, point cloud decoding method, and relevant apparatuses

Provided are a point cloud encoding method and decoding method, and an encoder and a decoder. The encoding method comprises: processing attribute information of a target point in a point cloud, so as to obtain a predicted residual of the attribute information of the target point; quantizing the predicted residual on the basis of a quantized weight of the target point and a quantized stride of the target point, so as to obtain a quantized residual of the attribute information of the target point, wherein the quantized weight of the target point is the weight used when weighting the quantized stride of the target point; and encoding the quantized residual, so as to obtain a code stream.

Point cloud encoding method and decoding method, and encoder and decoder

A method for detecting and tracking target object in a captured video using convolutional neural network (CNN) is provided. The method includes: inputting image data into a detecting model to generate detection results, wherein the detecting model is constructed by the CNN; inputting the image data into tracking models to generate tracking results; performing detection score enhancement operation according to the detection results and the tracking results to obtain enhanced detection results; matching the enhanced detection results and the tracking results by a matching operation; processing matched results and unmatched target detection results and unmatched target tracking results; and selectively updating the tracking models using tracking reliability estimation according to the matched results.

Method, device, and system for detecting and tracking objects in captured video using convolutional neural network

The present application discloses a cloud-network integration (CNI) oriented multi-access edge computing (MEC) architecture. An access network (AN) side of the architecture is provided with a plurality of edge computing nodes, a physical channel of the AN side is split into a plurality of subchannels with each of the subchannels supporting a media access control (MAC) access mode, and a software defined network (SDN) controller is arranged in the architecture and is configured to allocate resources of the subchannels at a physical layer and protocols at an MAC layer and control offloading at the edge computing nodes or a cloud center by a terminal user. The present application can carry out fine-grained control and cooperative management on network resources and computing resources, and achieve more effective computation offloading and service enhancement.

Cloud-network integration oriented multi-access edge computing architecture

A braking method includes: obtaining a first state information of the vehicle, which includes a vehicle mass and a deceleration required by braking; calculating a braking torque according to the first state information, and controlling the vehicle to output an electric braking torque according to the braking torque; obtaining a current vehicle speed and a mechanical braking application delay time; calculating an electric braking exit speed according to the braking torque required by the vehicle and the deceleration required by braking; calculating a mechanical braking application speed according to the mechanical braking application delay time, the deceleration required by braking, and the electric braking exit speed; and determining whether to control the vehicle to unload the electric braking torque, and whether to control the vehicle to apply a mechanical braking torque according to the current vehicle speed, the electric braking exit speed, and the mechanical braking application speed.

Vehicle and braking method and device therefor

A driving control method and apparatus for a vehicle, and a vehicle are provided, and relate to the field of vehicle control. The vehicle includes at least two carriages, and the driving control method includes the following steps: obtaining battery level information of a power battery corresponding to each of the carriages; obtaining at least one of a level allocated to each carriage or a payload allocated to each carriage according to the battery level information of the power battery corresponding to each carriage; and obtaining an output torque of each carriage according to the at least one of the level allocated to each carriage or the payload allocated to each carriage.

Method for balancing batteries of a vehicle, and device, and vehicle

A multi-functional, computer-aided gastroscopy system optimized with integrated AI solutions is disclosed. The system makes use of multiple deep-learning neural models to achieve low latency and high-performance requirements for multiple tasks. The optimization is made at three levels: architectural, modular and functional level. At architectural level, the models are designed in such a way that it is able to accomplish HP infection classification and detection of some lesions for one inference in order to reduce computation costs. At modular level, as a sub-model of HP infection classification, the site recognition model is optimized with temporal information. It not only improves the performance of HP infection classification, but also plays important roles for lesion detection and procedure status determination. At functional level, the inference latency is minimized by configuration and resource aware optimization. Also at functional level, the preprocessing is speeded up by image resizing parallelization and unified preprocessing.

Multi-functional computer-aided gastroscopy system optimized with integrated AI solutions and method

A braking method for a vehicle is provided. The method includes the following steps: obtaining a first state information of the vehicle, where the first state information includes a vehicle mass and a deceleration required by braking; calculating a braking torque required by the vehicle according to the first state information, and controlling an output of an electric braking torque according to the braking torque required by the vehicle; obtaining a current vehicle speed of the vehicle and a maximum electric braking exit speed; and; controlling, if the deceleration required by braking of the vehicle changes to zero, the vehicle to unload the electric braking torque when the current vehicle speed is less than the maximum electric braking exit speed. A braking device for a vehicle and a vehicle are further provided.

A control method for an electronic device and an electronic device are provided. The electronic device includes an infrared emitter, an infrared sensor, and a visible light sensor. The control method includes: acquiring an original pixel position of a zero-level region on the infrared sensor; acquiring a human eye pixel position of a human eye region on the visible light sensor; determining whether a human eye enters the zero-level region according to the original pixel position and the human eye pixel position; and triggering a protection mechanism of the infrared emitter when the human eye enters the zero-level region.

Control method for electronic device, electronic device and computer readable storage medium

A braking method for a vehicle is provided. The method includes the following steps: obtaining a first state information of the vehicle, where the first state information includes a vehicle mass and a deceleration required by braking; calculating a braking torque required by the vehicle according to the first state information, and controlling an output of an electric braking torque according to the braking torque required by the vehicle; obtaining a current gradient and a current vehicle speed of the vehicle; and determining whether to control the vehicle to unload the electric braking torque, and whether to control the vehicle to apply a mechanical braking torque according to the current vehicle speed, the braking torque required by the vehicle, the deceleration required by braking, and the current gradient. A braking device for a vehicle and a vehicle are further provided.

A vehicle and a braking method and a device therefor are provided. The method includes the following steps: obtaining a first state information of the vehicle, where the first state information includes a vehicle mass and a deceleration required by braking; calculating a braking torque required by the vehicle according to the first state information, and controlling the vehicle to output an electric braking torque according to the braking torque required by the vehicle; obtaining a current vehicle speed of the vehicle and an electric braking exit protection speed; and calculating an electric braking exit speed according to the braking torque required by the vehicle and the deceleration required by braking, and controlling the vehicle to unload the electric braking torque when the current vehicle speed is less than a larger one of the electric braking exit speed and the electric braking exit protection speed.

A projection module and a terminal are provided. The projection module includes a base, a housing, a first light source, a second light source and an optical element. The housing is disposed on the base, and defines an accommodating cavity together with the base. The first light source is disposed on the base and arranged in the accommodating cavity. The second light source is disposed on the base and arranged in the accommodating cavity. The optical element is disposed on the housing and includes a diffraction area and a diffusion area. The first light source aligns with the diffraction area, the second light source aligns with the diffusion area, the diffraction area is configured to diffract light passing through the diffraction area, and the diffusion area is configured to diffuse light passing through the diffusion area.

Projection module and terminal

Growing community of inventors

Shenzhen, China

Lu Wang