Abstract: Methods are disclosed for producing highly purified recombinant human insulin (RHI) having a purity of 99.0% (w/w) or greater, a Total Impurity (not including the related substance desamido AsnA21-RHI, as specified by USP) of 0.8% (w/w) or less, and an impurity C of 0.1% (w/w) or less. Also disclosed are API compositions of highly purified RHI having a purity of 99.0% (w/w) or greater, a Total Impurity of 0.8% (w/w) or less, and an impurity C of 0.1% (w/w) or less.

Abstract: Disclosed is a method for generating a semantic image labeling model, comprising: forming a first CNN and a second CNN, respectively; randomly initializing the first CNN; inputting a raw image and predetermined label ground truth annotations to the first CNN to iteratively update weights thereof so that a category label probability for the image, which is output from the first CNN, approaches the predetermined label ground truth annotations; randomly initializing the second CNN; inputting the category label probability to the second CNN to correct the input category label probability so as to determine classification errors of the category label probabilities; updating the second CNN by back-propagating the classification errors; concatenating the updated first and second CNNs; classifying each pixel in the raw image into one of general object categories; and back-propagating classification errors through the concatenated CNN to update weights thereof until the classification errors less than a predetermined

Abstract: A semantic segmentation model training method includes: performing, by a semantic segmentation model, image semantic segmentation on at least one unlabeled image to obtain a preliminary semantic segmentation result as the category of the unlabeled image; obtaining, by a convolutional neural network based on the category of the at least one unlabeled image and the category of at least one labeled image, sub-images respectively corresponding to the at least two images and features corresponding to the sub-images, where the at least two images comprise the at least one unlabeled image and the at least one labeled image, and the at least two sub-images carry the categories of the corresponding images; and training the semantic segmentation model on the basis of the categories of the at least two sub-images and feature distances between the at least two sub-images.

Abstract: The application relates to an array substrate, first metal line, second metal line and common electrode line insulated from each other and stacked in the fan-out area of the array substrate, on any cross section perpendicular to the extension path of first metal line, first metal line comprises a first and a second end in first direction, second metal line does not exceed first end in the first direction, and common electrode line does not exceed second end in the first direction. A staggered and stacked structure formed in the first direction by the second metal line and common electrode line can avoid electric conduction between second metal line and common electrode line in the case of poor cutting and ensure no short circuits occur for the array substrate. The application further relates to a liquid crystal display and an electronic device equipped with above array substrate.

Abstract: A method for identifying social relation of persons in an image, including: generating face regions for faces of the persons in the image; determining at least one spatial cue for each of the faces; extracting features related to social relation for each face from the face regions; determining a shared facial feature from the extracted features and the determined spatial cue, the determined feature being shared by multiple the social relation inferences; and predicting the social relation of the persons from the shared facial feature.

Abstract: The application relates to an array substrate, first metal line, second metal line and common electrode line insulated from each other and stacked in the fan-out area of the array substrate, on any cross section perpendicular to the extension path of first metal line, first metal line comprises a first and a second end in first direction, second metal line does not exceed first end in the first direction, and common electrode line does not exceed second end in the first direction. A staggered and stacked structure formed in the first direction by the second metal line and common electrode line can avoid electric conduction between second metal line and common electrode line in the case of poor cutting and ensure no short circuits occur for the array substrate. The application further relates to a liquid crystal display and an electronic device equipped with above array substrate.

Abstract: A chat message processing method is provided for a server. The method includes receiving a chat message sent by a sending terminal, where the chat message carries information about a receiving terminal, and forwarding the chat message to the receiving terminal. The method also includes, when determining that the chat message comprises an image, generating at least one group of descriptive texts used for describing the image; and sending the at least one group of descriptive texts of the image to the receiving terminal to cause the receiving terminal to, when displaying the image, simultaneously display the at least one group of descriptive texts as at least one candidate entering-item associated with the image.

Abstract: A method and an apparatus for segmenting a video object, an electronic device, a storage medium, and a program include: performing, among at least some frames of a video, inter-frame transfer of an object segmentation result of a reference frame in sequence from the reference frame, to obtain an object segmentation result of at least one other frame among the at least some frames; determining other frames having lost objects with respect to the object segmentation result of the reference frame among the at least some frames; using the determined other frames as target frames to segment the lost objects, so as to update the object segmentation results of the target frames; and transferring the updated object segmentation results of the target frames to the at least one other frame in the video in sequence. The accuracy of video object segmentation results can therefore be improved.

Abstract: A method for determining a candidate input is provided. Text that precedes and/or follows a current text input location in an input interface is acquired. A candidate sentence is acquired based on the text which is contained in a current sentence that corresponds to the current text input location. A preset number of possible conjunctions for the candidate sentence are determined based on occurrence probabilities of the possible conjunctions that are determined for the candidate sentence. The preset number of possible conjunctions are connected to the candidate sentence to obtain predicted candidate sentences. Occurrence probabilities of the predicted candidate sentences are calculated according to the text that precedes and/or follows the current text input location. Further, a preset number of the predicted candidate sentences are provided based on the calculated occurrence probabilities of the predicted candidate sentences as candidate inputs.

Abstract: The present invention relates to a linear regulator with real-time frequency compensation function, which belongs to the technical field of analog integrated circuits. The part of frequency compensation of the present invention includes the dual-frequency compensation networks and compensation transfer switcher, and the pulse delay circuit. The dual-frequency compensation networks and compensation transfer switcher provide the corresponding frequency compensation for linear regulator under two different capacitive loads. The pulse delay circuit generates a set of signals which have a delay related to the switch-pulse signals to control the access of the compensation transfer switcher and capacitive load.

Abstract: Disclosed is a method for manufacturing an aluminum structure having wood grain pattern is disclosed. The method includes cleaning a surface of the aluminum structure, imprinting a wood grain pattern on the cleaned surface of the aluminum structure, immersing the printed aluminum structure in an acidic solution to form an etched wood grain pattern along the imprinted wood grain pattern. The method further includes coating the surface along with the etched wood grain pattern with a polymeric powder and heating the polymeric powder coating on the surface of the aluminum structure. Thereafter, the method includes applying a heat transfer ink paper on a layer of heated polymeric powder coating, and heating the heat transfer ink paper for diffusing ink from the ink paper into at least a portion of the layer of heated polymeric powder coating to form the aluminum structure with wood grain pattern.

Abstract: Disclosed herein are a system and method for verifying face images based on canonical images. The method includes: retrieving, from a plurality of face images of an identity, a face image with a smallest frontal measurement value as a representative image of the identity; determining parameters of an image reconstruction network based on mappings between the retrieved representative image and the plurality of face images of the identity; reconstructing, by the image reconstruction network with the determined parameters, at least two input face images into corresponding canonical images respectively; and comparing the reconstructed canonical images to verify whether they belong to a same identity, where the representative image is a frontal image and the frontal measurement value represents symmetry of each face image and sharpness of the image. Thus, canonical face images can be reconstructed using only 2D information from face images under an arbitrary pose and lighting condition.

Abstract: The present invention relates to a linear regulator with real-time frequency compensation function, which belongs to the technical field of analog integrated circuits. The part of frequency compensation of the present invention includes the dual-frequency compensation networks and compensation transfer switcher, and the pulse delay circuit. The dual-frequency compensation networks and compensation transfer switcher provide the corresponding frequency compensation for linear regulator under two different capacitive loads. The pulse delay circuit generates a set of signals which have a delay related to the switch-pulse signals to control the access of the compensation transfer switcher and capacitive load.

Abstract: A method for identifying social relation of persons in an image, including: generating face regions for faces of the persons in the image; determining at least one spatial cue for each of the faces; extracting features related to social relation for each face from the face regions; determining a shared facial feature from the extracted features and the determined spatial cue, the determined feature being shared by multiple the social relation inferences; and predicting the social relation of the persons from the shared facial feature.

Abstract: Disclosed is a method for generating a semantic image labeling model, comprising: forming a first CNN and a second CNN, respectively; randomly initializing the first CNN; inputting a raw image and predetermined label ground truth annotations to the first CNN to iteratively update weights thereof so that a category label probability for the image, which is output from the first CNN, approaches the predetermined label ground truth annotations; randomly initializing the second CNN; inputting the category label probability to the second CNN to correct the input category label probability so as to determine classification errors of the category label probabilities; updating the second CNN by back-propagating the classification errors; concatenating the updated first and second CNNs; classifying each pixel in the raw image into one of general object categories; and back-propagating classification errors through the concatenated CNN to update weights thereof until the classification errors less than a predetermined

Abstract: A method and a system for exacting face features from data of face images have been disclosed. The system may comprise: A first feature extraction unit configured to filter the data of face images into a first plurality of channels of feature maps with a first dimension and down-sample the feature maps into a second dimension of feature maps; a second feature extraction unit configured to filter the second dimension of feature maps into a second plurality of channels of feature maps with a second dimension, and to down-sample the second plurality of channels feature maps into a third dimension of feature maps; and a third feature extraction unit configured to filter the third dimension of feature maps so as to further reduce high responses outside the face region such that reduce intra-identity variances of face images, while maintain discrimination between identities of the face images.

Abstract: An example method of pattern mining includes identifying, via a logic circuit, a plurality of patterns present in a plurality of transactions, the transactions including sets of items, and selecting, via the logic circuit, one of the identified patterns based on an occupancy of the identified pattern in the transactions.

Abstract: Disclosed herein are a system and method for verifying face images based on canonical images. The method includes: retrieving, from a plurality of face images of an identity, a face image with a smallest frontal measurement value as a representative image of the identity; determining parameters of an image reconstruction network based on mappings between the retrieved representative image and the plurality of face images of the identity; reconstructing, by the image reconstruction network with the determined parameters, at least two input face images into corresponding canonical images respectively; and comparing the reconstructed canonical images to verify whether they belong to a same identity, where the representative image is a frontal image and the frontal measurement value represents symmetry of each face image and sharpness of the image. Thus, canonical face images can be reconstructed using only 2D information from face images under an arbitrary pose and lighting condition.