Abstract: A method for product fault root cause analysis based on knowledge graphs includes (i) acquiring product-associated design data, wherein the design data comprises semi-structured data and/or unstructured data, (ii) acquiring product-associated production data, wherein the production data comprises structured data, (iii) performing information extraction on the design data to acquire structured design data, (iv) performing a knowledge fusion of knowledge generated separately based on the production data and the structured design data to construct a target knowledge graph, and (v) and performing product fault root cause analysis tasks based on the target knowledge graph.

Abstract: Provided is a method for moderating live streaming. The method includes: performing preliminary violation moderation on a current live streaming frame in a to-be-moderated live streaming room based on a high-accuracy moderation model and a high-recall moderation model that are cascaded with each other; and acquiring a target violation score of the current live streaming frame by inputting a first violation score of the current live streaming frame under the high-accuracy moderation model, a second violation score of the current live streaming frame under the high-recall moderation model, and a multi-dimensional behavior feature of the current live streaming frame in the to-be-moderated live streaming room to a pre-constructed behavior moderation model in response to the current live streaming frame passing the preliminary violation moderation.

Abstract: Provided is a human posture detection method including: acquiring a plurality of frames of image data; acquiring a plurality of human posture reference maps output by a human posture detection model responsive to inputting a current frame of image data to the human posture detection model with reference to human posture confidence maps of a previous frame of image data, wherein different human posture reference maps correspond different human-posture key points; identifying a human-posture key point in each of the human posture reference maps; and generating human posture confidence maps of the current frame of image data based on credibility of the human-posture key points, wherein the human posture confidence maps of the current frame of image data is configured to participate in generation of human posture confidence maps of a next frame of image data.

Abstract: Provided is a method for moderating live streaming. The method includes: performing preliminary violation moderation on a current live streaming frame in a to-be-moderated live streaming room based on a high-accuracy moderation model and a high-recall moderation model that are cascaded with each other; and acquiring a target violation score of the current live streaming frame by inputting a first violation score of the current live streaming frame under the high-accuracy moderation model, a second violation score of the current live streaming frame under the high-recall moderation model, and a multi-dimensional behavior feature of the current live streaming frame in the to-be-moderated live streaming room to a pre-constructed behavior moderation model in response to the current live streaming frame passing the preliminary violation moderation.

Abstract: Provided is a method for moderating abnormal users. The method includes: acquiring history behavior data of a plurality of to-be-moderated users, wherein the history behavior data includes behavior data of interactions of the plurality of to-be-moderated users; extracting a plurality of predetermined valid features from the history behavior data for each of the plurality of to-be-moderated users, wherein the predetermined valid features are features predetermined in sample data; calculating, based on predetermined probabilities of events associated with the plurality of valid features, a probability that each of the plurality of to-be-moderated users is the abnormal user; establishing a total probability function based on probabilities that the plurality of to-be-moderated users are the abnormal users; determining candidate users by solving, based on predetermined conditions, a maximum value of the total probability function; and determining the abnormal users by moderating the candidate users.

Abstract: Provided is a human posture detection method including: acquiring a plurality of frames of image data; acquiring a plurality of human posture reference maps output by a human posture detection model responsive to inputting a current frame of image data to the human posture detection model with reference to human posture confidence maps of a previous frame of image data, wherein different human posture reference maps correspond different human-posture key points; identifying a human-posture key point in each of the human posture reference maps; and generating human posture confidence maps of the current frame of image data based on credibility of the human-posture key points, wherein the human posture confidence maps of the current frame of image data is configured to participate in generation of human posture confidence maps of a next frame of image data.

Abstract: A multilayer mirror (M) reflecting extreme ultraviolet (EUV) radiation from a first wavelength range in an EUV spectral region includes a substrate (SUB) and a stack of layers (SL). The stack of layers has layers having a low index material and layers having a high index material. The low index material has a lower real part of the refractive index than does the high index material at a given operating wavelength in the first wavelength range. The stack of layers also includes a spectral purity filter on the stack of layers. The spectral purity filter is effective as an anti-reflection layer for ultraviolet (UV) radiation from a second wavelength range in a UV spectral region to increase an EUV-UV-reflectivity ratio of the multilayer mirror. The spectral purity filter (SPF) includes a non-diffractive graded-index anti-reflection layer (GI-AR) effective to reduce reflectivity in the second wavelength range.