Tsinghua University

China Mobile Online Services Company Limited

Nantong Advanced Communication Technology Research Institute Co., Ltd.

Yiping Duan

Xiaoming Tao

Ziqi Zhao

Chaoyi Han

Ning Ge

Jianhua Lu

Bingrui Geng

Qiwei Song

Qing Chen

Mingzhe Li

Danlan Huang

Dingcheng Gao

A method includes: pre-processing initial data of an acquired to-be-analyzed confrontation scenario, to obtain triplet data marked in a graph data structure, inputting the triplet data into a pre-trained analyzing model, to determine a result of analysis on semantic information of the to-be-analyzed confrontation scenario; wherein the analyzing model includes a plurality of operator networks that are provided with execution priorities, wherein the plurality of operator networks are configured for analyzing relations of different types; and the processing result of any one of the operator networks includes: in the relation type corresponding to the operator network, respective true relations of all of the node pairs, and confidences and descriptive values of the true relations; and according to the graph data structure that is updated by using the processing results of all of the operator networks, characterizing the result of analysis on the semantic information of the to-be-analyzed confrontation scenario.

Method and apparatus for semantic analysis on confrontation scenario based on target-attribute-relation

The disclosure provides a method and an apparatus for determining a quality grade of video data, and relates to the field of data processing technologies, wherein the method includes: acquiring a plurality of initial EEG data; based on the plurality of initial EEG data, determining an initial EEG data set, wherein the initial EEG data set includes a first sub-data set and a second sub-data set, the first sub-data set is a data set built on the basis of emotional response electroencephalogram data, and the second sub-data set is a data set built on the basis of electroencephalogram emotion data; processing the first sub-data set and the second sub-data set by using a transfer learning algorithm to obtain a third sub-data set and a fourth sub-data set; and based on the third sub-data set and the fourth sub-data set, determining a quality evaluation grade of video data with degraded quality.

Method and apparatus for determining quality grade of video data

A semantic representation method and system based on an aerial surveillance video, and an electronic device are provided. The semantic representation method includes: taking a pedestrian and a vehicle in the aerial surveillance video as a target for tracking; inputting a coordinate track of the target into a first semantic classifier to output a first semantic result of the target; performing semantic merging on the first semantic result, and inputting an obtained semantic merging result into a second semantic classifier to output a second semantic result of the target; and performing cluster analysis on the first semantic result to obtain a target group of the target, and according to the target group of the target, the obtained scene analysis result and the second semantic result, determining semantics of the aerial surveillance video.

Semantic representation method and system based on aerial surveillance video and electronic device

A variable bit rate generative compression method based on adversarial learning is provided. According to the method, a variance of a feature map of an encoding-decoding fill convolutional network is quantized to train a single generative model to perform variable bit rate compression. The method includes the following implementation steps of: constructing training and testing data sets through an image acquisition device; constructing a generative compression network based on an auto-encoder structure; according to a rate-distortion error calculation unit, alternately training a generative network; according to a target compression rate, calculating a mask threshold; based on a feature map channel redundancy index, calculating a mask; and performing lossless compression and decoding on the mask and the feature map. According to the invention, only a single model is trained, but compression results with different bit rates can be generated, and on a limit compression rate below 0.1 bpp.

Variable bit rate generative compression method based on adversarial learning

Growing community of inventors

Beijing, China

Yiping Duan