A character restoration method and apparatus, a storage medium, and an electronic device are provided. The character restoration method includes: a character identifier of a character in a text region is determined, where the character identifier is used for uniquely identifying the character; and encoding is performed at least according to the character identifier, and encoded data is sent to a receiving end, where the encoded data is used for the receiving end to decode the encoded data and restore the character according to the character identifier obtained after decoding, that is, encoding is performed merely according to a small amount of information, and then the information is obtained by decoding, so as to restore the character.

Technology is described herein for generating questions using a neural network. The technology generates the questions in a three-step process. In the first step, the technology selects, using a first neural network, a subset of textual passages from an identified electronic document. In the second step, the technology generates, using a second neural network, one or more candidate answers for each textual passage selected by the first neural network, to produce a plurality of candidate passage-answer pairs. In the third step, the technology selects, using a third neural network, a subset of the plurality of candidate passage-answer pairs. The technology then generates an output result that includes one or more output questions chosen from the candidate passage-answer pairs produced by the third neural network. The use of the first neural network reduces the processing burden placed on the second and third neural networks. It also reduces latency.

Technology is described herein for generating questions using a neural network. The technology generates the questions in a three-step process. In the first step, the technology selects, using a first neural network, a subset of textual passages from an identified electronic document. In the second step, the technology generates, using a second neural network, one or more candidate answers for each textual passage selected by the first neural network, to produce a plurality of candidate passage-answer pairs. In the third step, the technology selects, using a third neural network, a subset of the plurality of candidate passage-answer pairs. The technology then generates an output result that includes one or more output questions chosen from the candidate passage-answer pairs produced by the third neural network. The use of the first neural network reduces the processing burden placed on the second and third neural networks. It also reduces latency.

An example method is provided for encoding text for language processing. The method may be executed by a processing system, and the method includes receiving text comprising a plurality of alphanumeric characters or symbols and converting the text into a numerical vector comprising a plurality of numerical values, by mapping each alphanumeric character or symbol of the text to a vertex coordinate of one of a plurality of vertices of a hypercube, wherein a number of the plurality of vertices is equal to or greater than a number of the plurality of alphanumeric characters or symbols, wherein the numerical vector consumes less space in memory than the text. An amount of time consumed by language processing of the numerical vector may be less than an amount of time consumed by language processing of the text.

An example method is provided for encoding text for language processing. The method may be executed by a processing system, and the method includes receiving text comprising a plurality of alphanumeric characters or symbols and converting the text into a numerical vector comprising a plurality of numerical values, by mapping each alphanumeric character or symbol of the text to a vertex coordinate of one of a plurality of vertices of a hypercube, wherein a number of the plurality of vertices is equal to or greater than a number of the plurality of alphanumeric characters or symbols, wherein the numerical vector consumes less space in memory than the text. An amount of time consumed by language processing of the numerical vector may be less than an amount of time consumed by language processing of the text.

Examples are provided for pre-training a computer vision foundation model. A representative method comprises curating a pre-training database of image-text pairs from weakly labeled data. Language is encoded of text descriptions from the image-text pairs. The images of the image-text pairs are encoded using a hierarchical vision transformer with shifted windows and convolutional embedding. Based on the encoded images and the encoded language, the computer vision foundation model is pre-trained via unified image-text contrastive learning.

Examples are provided for pre-training a computer vision foundation model. A representative method comprises curating a pre-training database of image-text pairs from weakly labeled data. Language is encoded of text descriptions from the image-text pairs. The images of the image-text pairs are encoded using a hierarchical vision transformer with shifted windows and convolutional embedding. Based on the encoded images and the encoded language, the computer vision foundation model is pre-trained via unified image-text contrastive learning.

Embodiments described herein a CROss-Modal Distribution Alignment (CROMDA) model for vision-language pretraining, which can be used for retrieval downstream tasks. In the CROMDA mode, global cross-modal representations are aligned on each unimodality. Specifically, a uni-modal global similarity between an image/text and the image/text feature queue are computed. A softmax-normalized distribution is then generated based on the computed similarity. The distribution thus takes advantage of property of the global structure of the queue. CROMDA then aligns the two distributions and learns a modal invariant global representation. In this way, CROMDA is able to obtain invariant property in each modality, where images with similar text representations should be similar and vice versa.

Embodiments described herein a CROss-Modal Distribution Alignment (CROMDA) model for vision-language pretraining, which can be used for retrieval downstream tasks. In the CROMDA mode, global cross-modal representations are aligned on each unimodality. Specifically, a uni-modal global similarity between an image/text and the image/text feature queue are computed. A softmax-normalized distribution is then generated based on the computed similarity. The distribution thus takes advantage of property of the global structure of the queue. CROMDA then aligns the two distributions and learns a modal invariant global representation. In this way, CROMDA is able to obtain invariant property in each modality, where images with similar text representations should be similar and vice versa.

A method for supporting block coding is provided. The method includes the steps of: determining an arrangement position of a coding block selected by a user on the basis of a sentence component-specific arrangement order specified from a sentence structure of a spoken language; and providing the user with an arrangement result of the coding block specified on the basis of the arrangement position.