Systems, methods and products for classifying images according to a visual concept where, in one embodiment, a system includes an object detector and a visual concept classifier, the object detector being configured to detect objects depicted in an image and generate a corresponding object data set identifying the objects and containing information associated with each of the objects, the visual concept classifier being configured to examine the object data set generated by the object detector, detect combinations of the information in the object data set that are high-precision indicators of the designated visual concept being contained in the image, generate a classification for the object data set with respect to the designated visual concept, and associate the classification with the image, wherein the classification identifies the image as either containing the designated visual concept or not containing the designated visual concept.

Systems, methods and products for classifying images according to a visual concept where, in one embodiment, a system includes an object detector and a visual concept classifier, the object detector being configured to detect objects depicted in an image and generate a corresponding object data set identifying the objects and containing information associated with each of the objects, the visual concept classifier being configured to examine the object data set generated by the object detector, detect combinations of the information in the object data set that are high-precision indicators of the designated visual concept being contained in the image, generate a classification for the object data set with respect to the designated visual concept, and associate the classification with the image, wherein the classification identifies the image as either containing the designated visual concept or not containing the designated visual concept.

A method for training a text positioning model includes: obtaining a sample image, where the sample image contains a sample text to be positioned and a text marking box for the sample text; inputting the sample image into a text positioning model to be trained to position the sample text, and outputting a prediction text box for the sample image; obtaining a sample prior anchor box corresponding to the sample image; and adjusting model parameters of the text positioning model based on the sample prior anchor box, the text marking box and the prediction text box, and continuing training the adjusted text positioning model based on a next sample image until model training is completed, to generate a target text positioning model.

Embodiments described herein provide a computer-implemented method of creating a digest of a document. The document to be processed and analysed may be a physical document, or it may already be in a digital form. In the case of starting from a physical document, the document is first scanned, so as to obtain an image of the document. The digital document is then processed using an algorithm or function to obtain one or more datasets comprising a plurality of position independent values. Each of the datasets may correspond to a different line of text or field of text within the document. The one or more datasets are then encoded, the encoded data being used to generate a digest associated with the document, and wherein the digest comprises a plurality of short hashes corresponding to each dataset. The generated digest can then be used to print a digital signature on the document, which can be used to later verify the authenticity of the document or a copy thereof.

The present disclosure relates to a method of evaluating accuracy of a summary of a document. The method includes receiving a plurality of reference summaries of a document and a system summary of the document. The system summary is generated by a machine. The method further includes extracting, for each reference summary, a tuple that is a pair of words composed of a modified word and a dependent word having a dependency relation to the modified word and a label representing the dependency relation. The method further includes replacing, for each of the extracted tuples, each of the modified word of the tuple's word pair and the dependent word with a class predetermined for the words. The method further generates a score of the system summary based on the class and a set of tuples of the system summary.

The disclosure provides a calculation practicing method, a system, an electronic device and a computer readable storage medium, the calculation practicing method includes: providing a calculation question; identifying the type and content of the calculation question; generating an answer area according to the type and content of the calculation question; receiving an answering operation in which the user inputs the answer string for the calculation question in the answer area; identifying the answer string inputted by the user; and determining whether each of the answer characters in the answer string is correct, if there is an incorrect answer character, it will be marked, so that the calculation practice can be realized through the electronic device, which is convenient for students to carry out training.

Methods and systems for determining relevance for a new document are described. Existing documents that have a high probability of relevance can be chosen. A vocabulary of words in the existing documents can be built. Each word can be mapped into a vector such that each existing document can be represented by a sequence of vectors and each sentence and/or paragraph in each existing document can be represented by a subsequence of vectors including a subset of the sequence of vectors. Data augmentation can be applied changing an order of the subsequences in order to create additional documents represented by the subsequences. A deep neural network can be trained using the subsequences that represent the existing documents and the subsequences that represent additional documents. The new documents can be trained using a trained deep neural network. A relevant document can be output using the trained deep neural network.

Methods and systems for determining relevance for a new document are described. Existing documents that have a high probability of relevance can be chosen. A vocabulary of words in the existing documents can be built. Each word can be mapped into a vector such that each existing document can be represented by a sequence of vectors and each sentence and/or paragraph in each existing document can be represented by a subsequence of vectors including a subset of the sequence of vectors. Data augmentation can be applied changing an order of the subsequences in order to create additional documents represented by the subsequences. A deep neural network can be trained using the subsequences that represent the existing documents and the subsequences that represent additional documents. The new documents can be trained using a trained deep neural network. A relevant document can be output using the trained deep neural network.

The present disclosure relates to utilizing a deep recurrent neural network for accurately performing punctuation restoration. For example, the disclosed systems can provide a sequence of words to a punctuation restoration neural network having multiple bi-directional recurrent layers and one or more neural attention mechanisms. In one or more embodiments, the punctuation restoration neural network incorporates layer-wise attentions and/or multi-head attention. The disclosed systems can utilize the punctuation restoration neural network to generate probabilities for each word, indicating the likelihood that each possible punctuation mark is associated with that word. Based on these probabilities, the disclosed systems can generate a punctuated transcript that includes punctuation before the appropriate words.

The disclosed embodiments describe methods, systems, and apparatuses for determining user intent. A method is disclosed comprising obtaining a session text of a user; calculating, by the processor, a feature vector based on the session text; determining probabilities that the session text belongs to a plurality of intent labels, the probabilities calculated using a multi-level hierarchal intent classification model, the intent labels assigned to levels in the multi-level hierarchal intent classification model; and assigning a user intent to the session text based on the probabilities.