Abstract: A method for generating a vector representation of a text includes dividing the text into text segments. Each text segment is represented as a segment vector corresponding to the respective text segment by employing a first-level semantic model. The segment vector is configured to indicate a semantics of the text segment. Text semantics recognition is performed on the segment vector of each text segment by employing a second-level semantic model to obtain a text vector for indicating a topic of the text.

Abstract: Aspects of the present disclosure involve a system and method for optimizing a content shared via a messaging system based on an expected audience for the content. A content is received from a publisher client device associated with a publisher. A list of content versions associated with the publisher is accessed form a database, each content version being associated with an associated set of content parameters. The one or more content versions of the content are generated. When an access request is received from a viewer client device, an optimal content version is identified for the viewer client device based on the device capabilities of the viewer client device. An updated list of content versions is associated with the publisher by aggregating the optimal content versions identified for a plurality of viewer client devices, and the database is updated with the updated list of content versions associated with the publisher.

Abstract: The present disclosure provides a method, apparatus, device and storage medium for training a dialogue understanding model, and relates to technical field of computers, and specifically to the technical field of artificial intelligence such as natural language processing and deep learning. The method for training a dialogue understanding model includes: obtaining dialogue understanding training data; performing joint training for a dialogue understanding pre-training task and a general pre-training task by using the dialogue understanding training data, to obtain a dialogue understanding model. According to the present disclosure, a model specially adapted for a dialogue understanding task may be obtained by training.

Abstract: The present disclosure provides a multi-lingual model training method, apparatus, electronic device and readable storage medium and relates to the technical field of deep learning and natural language processing. A technical solution of the present disclosure when training the multi-lingual model is: obtaining training corpuses comprising a plurality of bilingual corpuses and a plurality of monolingual corpuses; training a multi-lingual model with a first training task by using the plurality of bilingual corpuses; training the multi-lingual model with a second training task by using the plurality of monolingual corpuses; and completing the training of the multi-lingual model in a case of determining that loss functions of the first training task and second training task converge.

Abstract: A universal wireless charging receiver for an electronic device is provided. The universal wireless charging receiver includes a base casing, a first magnetic shielding element, a power supply module and an induction coil. The first magnetic shielding element, the power supply module and the induction coil are installed in the base casing. The power supply module is located over the first magnetic shielding element. A magnetic field is shielded by the first magnetic shielding element. Consequently, an eddy current is not generated by the power supply module through the magnetic field.

Abstract: Aspects of the present disclosure involve a system and method for providing the best available version of a content to a viewer client device considering the device capabilities of the viewer client device. To access the content, a viewer client device sends an access request comprising its device capabilities. Upon receiving the access request, the server computer generates a ranked list of all possible content versions for the viewer client device, the content versions being ranked based on their compatibility with the device capabilities of the viewer client device. The server computer selects, among the one or more available versions of the content, the content version ranked first in the ranked list and provides the selected content version to the viewer client device.

Abstract: A method for correcting an error in a text, an electronic device, and a storage medium are provided. The method includes: obtaining an original text; obtaining a training text by preprocessing the original text; extracting a plurality of feature vectors corresponding to each word in the training text; obtaining an input vector by processing the plurality of feature vectors; obtaining a target text by inputting the input vector into a text error correction model; and adjusting parameters of the text error correction model based on a difference between the target text and the original text.

Abstract: The subject technology determines, by at least one processor, that a value of a capability parameter is not available for a first recipient device, the first recipient device being one of a plurality of recipient devices. The subject technology receives, from the first recipient device, a value of the capability parameter. The subject technology clusters, by at least one processor, the plurality of recipient devices into a plurality of groups of recipient devices based on device capabilities including an available storage space. The subject technology determines, for each group of recipient devices, a set of content parameters compatible with device capabilities of the plurality of recipient devices of a group of client devices.

Abstract: A method, an apparatus, a device and a storage medium for learning a knowledge representation are provided. The method can include: sampling a sub-graph of a knowledge graph from a knowledge base; serializing the sub-graph of the knowledge graph to obtain a serialized text; and reading using a pre-trained language model the serialized text in an order in the sub-graph of the knowledge graph, to perform learning to obtain a knowledge representation of each word in the serialized text. The knowledge representation learning in this embodiment is performed for entity and relationship representation learning in the knowledge base.

Abstract: Among other things, embodiments of the present disclosure improve the functionality of electronic messaging and imaging software and systems by automating the client-side transcoding of video data based on content. For example, an appropriate transcoding configuration can be selected for video data having complex motion or textures. Accordingly, video quality can be improve when complex motions or textures are present.

Abstract: A method for resource sorting, a method for training a sorting model and corresponding apparatuses which relate to the technical field of natural language processing under artificial intelligence are disclosed. The method according to some embodiments includes: forming an input sequence in order with an item to be matched and information of candidate resources; performing Embedding processing on each Token in the input sequence, the Embedding processing including: word Embedding, position Embedding and statement Embedding; and inputting result of the Embedding processing in a sorting model to obtain sorting scores of the sorting model for the candidate resources, the sorting model is obtained by pre-training of a Transformer model.

Abstract: The subject technology accesses, by at least one processor, device capabilities for each of a plurality of recipient devices. The subject technology clusters, by the at least one processor, the plurality of recipient devices into a plurality of groups of recipient devices based on device capabilities. The subject technology determines, for each group of recipient devices, a set of content parameters compatible with the device capabilities of the plurality of recipient devices of a group of client devices. Further, the subject technology generates, for each group of recipient devices, a version of the content according to an associated set of content parameters.

Abstract: A method for training a pre-trained knowledge model includes: obtaining a training text, in which the training text includes a structured knowledge text and an article corresponding to the structured knowledge text, and the structured knowledge text includes a head node, a tail node, and a relationship between the head node and the tail node; and training a pre-trained knowledge model to be trained according to the training text.

Abstract: A system, automobile, and method for implementing an electronic control function of an automobile. The system includes a first vehicle integration unit (VIU), an automobile control unit, and a plurality of automobile parts. The automobile control unit includes a first domain controller (DC) or a central computing platform (CCP). The automobile control unit is configured to send first control information to the first VIU. The first VIU is configured to control the plurality of automobile parts based on the first control information. In embodiments of this application, the first VIU controls the plurality of automobile parts.

Abstract: A method for generating a vector representation of a text includes dividing the text into text segments. Each text segment is represented as a segment vector corresponding to the respective text segment by employing a first-level semantic model. The segment vector is configured to indicate a semantics of the text segment. Text semantics recognition is performed on the segment vector of each text segment by employing a second-level semantic model to obtain a text vector for indicating a topic of the text.

Abstract: Among other things, embodiments of the present disclosure improve the functionality of electronic messaging and imaging software and systems by automating the client-side transcoding of video data based on content. For example, an appropriate transcoding configuration can be selected for video data having complex motion or textures. Accordingly, video quality can be improve when complex motions or textures are present.

Abstract: The subject technology accesses, by at least one processor, device capabilities for each of a plurality of recipient devices. The subject technology clusters, by the at least one processor, the plurality of recipient devices into a plurality of groups of recipient devices based on device capabilities. The subject technology determines, for each group of recipient devices, a set of content parameters compatible with the device capabilities of the plurality of recipient devices of a group of client devices. Further, the subject technology generates, for each group of recipient devices, a version of the content according to an associated set of content parameters.

Abstract: An example method for encoding or decoding video data includes storing, by a video coder and in a reference picture buffer, a version of a current picture of the video data, including the current picture in a reference picture list (RPL) used to predict the current picture, and coding, by the video coder and based on the RPL, a block of video data in the current picture based on a predictor block of video data included in the version of the current picture stored in the reference picture buffer.

Abstract: Aspects of the present disclosure involve a system and method for performing operations comprising receiving, by one or more processors, from a sender device, a content to be shared with one or more recipients via a messaging application; identifying, by the one or more processors, at least one recipient device for each recipient; accessing, by the one or more processors, device capabilities for each of the one or more recipient devices; determining, by the one or more processors, a set of content parameters compatible with the device capabilities of the recipient devices; and generating, by the one or more processors, a version of the content according to the set of content parameters.

Abstract: The present invention discloses a live pig quantity checking apparatus and method. The checking method includes: obtaining first electronic ear tag numbers of electronic ear tags within a period of time interval threshold range; determining whether live pigs in a monitored piggery all have a food intake activity, and if yes, determining a quantity of the live pigs according to the first electronic ear tag numbers; and if no, determining live pigs not having the food intake activity according to the first electronic ear tag numbers. The live pig quantity checking apparatus and method of the present invention improves the live pig quantity checking efficiency.