Huawei Technologies Co., Limited

Peking University

Shen Zhang

Qikun Wei

Wen Gao

Ronggang Wang

Zhenyu Wang

Jie Xiong

Youqing Yang

Yihong Huang

Qiujie Fu

Yuping Jiang

Ruichuan You

The present invention provides a live video transmission method and system and an apparatus. According to the method, an M2U device receives a UDP request sent by a client, where the UDP request includes a live channel identifier and is used to request video data of a live channel corresponding to the live channel identifier, and the live channel identifier includes a multicast address or a URL. Then, the M2U device obtains the video data corresponding to the live channel identifier, and obtains a source IP address and a source port of the UDP request, where the source IP address and the source port of the UDP request are used as a destination IP address and a destination port of a UDP unicast packet to be subsequently sent to the client.

Live video transmission method and system, and apparatus

A hot live video determining method is provided. A network device first obtains a plurality of video request messages for a target video; then determines statistical information of the target video based on the plurality of video request messages, where the statistical information includes a quantity of video request messages for the target video that are obtained in specified duration and fragment information of a video fragment requested in each video request message; and finally determines, based on the statistical information, whether the target video is a hot live video. According to the method, the network device can identify the hot live video in the Internet, so the network device can carry the hot live video in a multicast manner, to avoid breakdown that is of a network and a video server and that is caused by transmitting the hot live video in a unicast method.

Hot live video determining method and device

Embodiments of the present disclosure provide a method for accelerating CDVS extraction process based on a GPGPU platform, wherein for the stages of feature detection and local descriptor computation of the CDVS extraction process, operation logics and parallelism strategies of respective inter-pixel parallelism sub-procedures and respective inter-feature point parallelism sub-procedures are implemented by leveraging an OpenCL general-purpose parallelism programming framework, and acceleration is achieved by leveraging a GPU's parallelism computation capability; including: partitioning computing tasks for a GPU and a CPU; reconstructing an image scale pyramid storage model; assigning parallelism strategies to respective sub-procedures for the GPU; and applying local memory to mitigate the access bottleneck. The technical solution of the present disclosure may accelerate the CDVS extraction process and significantly enhances the extraction performance.

Method for accelerating a CDVS extraction process based on a GPGPU platform

Growing community of inventors

Nanjing, China

Shen Zhang