484results about "Digital ink recognition" patented technology

The present disclosure generally relates to embodiments for video communication interface for managing content that is shared during a video communication session.

A character processing method and apparatus, an electronic device and a storage medium. The method includes: acquiring a first image comprising a to-be-processed character; training a target stroke order determination model by combining with a spatial attention mechanism and a channel attention mechanism; and inputting the first image into the target stroke order determination model trained in advance to obtain a target stroke order corresponding to the to-be-processed character.

The invention belongs to the field of intelligent man-machine interaction, and relates to an intelligent writing and correcting method and system.The method comprises the steps that a flexible pressure sensor matrix and a coordinate positioning system are used for sensing multi-dimensional data generated when an input pen writes, and a four-dimensional writing feature space is constructed; a three-level correction system is adopted, and logical reasoning verification is conducted on the answers through a deep semantic understanding model; the three-stage correction system sequentially comprises OCR font matching, semantic correction based on a large language model and intelligent correction based on a process reward model; a cognitive diagnosis model is introduced, and the answer hesitation degree is judged by analyzing answer indexes; fusing question answering information, utilizing a large model to identify knowledge points to which questions belong, and inferring knowledge mastering weak points according to question answering conditions; writing information is accurately collected through an intelligent writing capture technology based on multi-modal sensing fusion, an intelligent correction integrated method is built in combination with a cognitive diagnosis model, and efficient recognition and accurate correction of written content are achieved.

The invention provides a handwritten mathematical expression recognition model training method, a handwritten mathematical expression recognition method and a handwritten mathematical expression recognition device, a LaTeX sequence is automatically extracted and enhanced from an academic PDF file, a large-scale and high-reality handwritten formula image data set is generated through rendering and material conversion, and then an end-to-end recognition model is trained. According to the scheme, the core bottleneck of training data scarcity and insufficient diversity caused by high manual labeling cost is fundamentally solved; according to the method, the generalization ability and robustness of the model in a real scene are effectively improved by synthesizing a vivid handwritten image; and finally, the identification model realizes a remarkable breakthrough in performance when processing a complex and long sequence formula, and a set of complete and extensible engineering solution is provided for large-scale application of the technology.

The invention discloses a handwritten mathematical formula dynamic recognition method based on multi-modal feature fusion, which comprises the following steps of: 1, performing code conversion on stroke track sequence information contained in data by adopting an HSV color model to generate an HSV pseudo-color image, and fusing the HSV pseudo-color image with a data binary image to obtain a time sequence rich image; 2, feature extraction is performed on the time sequence rich image and the stroke track sequence through a convolutional neural network, image features and stroke features are obtained, the stroke features are enhanced by using the image features, and multi-modal features are obtained after the image features and the enhanced stroke features are fused; step 3, enhancing the input multi-modal features by adopting position coding through a decoder, and performing symbol prediction in combination with an attention mechanism and a multi-scale symbol counting module to generate a mathematical formula symbol sequence; and 4, designing rules, and correcting the generated mathematical formula symbol sequence. According to the invention, the recognition accuracy of confusable characters can be improved.

The invention relates to the technical field of information digitization, in particular to an archive digitization management method and system and a storage medium. The method comprises the following steps: acquiring original scanning image data of the science and technology handwritten archive; performing regional brightness compensation processing on the original scanning image data to obtain handwriting feature data, the handwriting feature data including handwriting pressure change feature data and ink diffusion feature data; and performing local projection analysis on the original scanning image data according to the handwriting feature data, and performing breakpoint connectivity analysis to obtain standardized character sequence data including character inclination angle data and stroke connection feature data. According to the method, the details of the handwritten document can be better restored through comprehensive analysis of the handwriting pressure and the ink mark diffusion characteristics, the coherence of strokes and the actual form of characters are recognized, and the accuracy of character recognition is remarkably improved.

The invention relates to the technical field of intelligent whiteboards, in particular to a whiteboard handwritten formula real-time identification method and system based on artificial intelligence. The method comprises the following steps: constructing a handwriting path track based on a touch point sequence, and delimiting a high-density handwriting region; performing symbol detection on the high-density handwriting region to position a potential symbol region, and extracting a stroke feature vector of the potential symbol region through a MobileNetV3-Small model; using the stroke feature vector to fuse the handwriting pressure and the inclination angle feature to construct a symbol relation graph; and performing dynamic planning on the initial LaTeX sequence through a CYK algorithm on the basis of node characteristics and a dynamic adjacency matrix in the symbol relation graph, and converting the symbol relation graph into a LaTeX expression. According to the method, dynamic threshold segmentation is adopted, handwriting pressure and dip angle information are combined, the sampling density and the NMS threshold value are dynamically adjusted, a high-density area detection result is reserved preferentially, the symbol detection accuracy is remarkably improved, and the problem that missing detection or false detection is prone to occurring in a dense writing area in a traditional method is solved.

A handwriting geometry recognition and calibration system by using neural network and mathematical feature includes: a pre-processor for pre-processing coordinate points of geometric figures from user's handwriting so as to get a plurality of sample points which expresses the geometric figures to be recognized; a neural network connected to the pre-processor for receiving the sample points of the geometric figure and recognizing the geometric figure so as to acquire a coarse class of the geometric figure; and an mathematical logic unit connected to the neural network for receiving recognition results from the neural network, including coarse classifications which are used in a secondary classification by using conventional mathematical recognition logics so as to determine an exact geometry shape of the geometric figure; then the geometric figure being calibrated so as to get a geometry with a regular shape.

The present disclosure provides a character recognition model training method and apparatus, a character recognition method and apparatus, a device and a medium, relating to the technical field of artificial intelligence, and specifically to the technical fields of deep learning, image processing and computer vision, which can be applied to scenarios such as character detection and recognition technology. The specific implementing solution is: partitioning an untagged training sample into at least two sub-sample images; dividing the at least two sub-sample images into a first training set and a second training set; where the first training set includes a first sub-sample image with a visible attribute, and the second training set includes a second sub-sample image with an invisible attribute; performing self-supervised training on a to-be-trained encoder by taking the second training set as a tag of the first training set, to obtain a target encoder.

This application discloses a system for signing medical communication documents based on digital handwriting. The service governance platform Spring Cloud microservice framework is used as the core architecture layer of the business system. The access layer provides an access interface to access the access medium containing the electronic signature collection device; the access layer implements the intermediate authentication of the application layer and each business system of the smart hospital, and provides a service authentication entrance; the application layer provides signing task management, communication document management, signing document verification, and evidence information management services for the medical communication documents generated by each business system; the service layer converts the documents to be signed and unifies the format, calls the corresponding signing terminal to collect the signing handwriting according to the signing content, signing order, and signing position required by the signing task management to generate a signature data packet and merge the telegram to form a signed document, encrypts the signing data and signed documents and stores them; the resource layer supports the service layer, application layer service operation and signing server cluster resource equipment.

The present invention provides a method, system and computing device for recognizing the stroke order of handwritten Chinese characters, the method comprising: obtaining a trajectory point sequence of a handwritten Chinese character to be recognized, and preprocessing the trajectory point sequence of the handwritten Chinese character to obtain a key point sequence of the handwritten Chinese character; extracting the stroke features of the handwritten Chinese character based on the key point sequence of the handwritten Chinese character, wherein the stroke features include basic features and semantic features; and using a trained stroke order recognition model to process the stroke features of the handwritten Chinese character together with the stroke features of the corresponding template Chinese character to obtain a stroke order recognition result. The present invention can obtain handwritten Chinese characters and perform real-time analysis, accurately recognize whether the stroke order of writing the handwritten Chinese character is correct or not, and feed back the recognition result to the user, which can not only greatly improve the recognition efficiency of the stroke order of handwritten Chinese characters, relieve the tutoring and teaching pressure of parents and teachers, but also assist the writer to develop the habit of writing Chinese characters according to the correct stroke order.

The embodiment of the present application provides a brush effect image generation method, an image editing method, a device and a storage medium, the method comprising: responding to an area selection operation on a display interface in a display screen, determining a pattern extraction area in the display interface; obtaining a first brush effect image according to the pattern extraction area, the pattern of the first brush effect image being the pattern corresponding to the pattern extraction area in the display interface; obtaining a second brush effect image according to the first brush effect image, wherein the patterns of the first brush effect image and the second brush effect image are the same, and the clarity of the second brush effect image is greater than that of the first brush effect image. The brush effect image obtained by using the embodiment of the present application has a higher clarity, which improves the user experience.

The invention discloses a handwritten digital image clustering method, system and device based on tensor anchor point label learning and a medium, and the method comprises the steps: sequentially carrying out the preprocessing, multi-view feature data generation, feature selection and feature extraction, and feature normalization processing of a to-be-clustered handwritten digital image, and obtaining multi-view handwritten digital image feature data; constructing a target function of handwritten digital image clustering based on the multi-view handwritten digital image feature data; introducing an auxiliary variable into the target function, and introducing a penalty term by adopting an augmented Lagrangian multiplier method to obtain a Lagrangian function; solving the Lagrangian function by adopting an alternating iteration method to obtain a clustering result of final division of the handwritten digital image; similarity information from different view angle features is mined through the anchor point technology, the quality of each view angle clustering result is controlled through tensor constraint, the problem that handwritten digital images cannot be efficiently processed through a traditional multi-view angle clustering method is effectively solved, and the method has the advantages of being high in clustering precision and high in efficiency.

The invention provides a handwritten signature recognition method and system fused with a multi-modal technology in the technical field of handwritten signature recognition, and the method comprises the steps: S1, obtaining a large number of handwritten signature images, carrying out the preprocessing and marking of each handwritten signature image, constructing a real data set, and carrying out the sample expansion operation of the real data set; s2, a large number of handwritten documents are obtained, and data synthesis operation is executed based on all the handwritten documents to construct a synthesis data set; s3, creating a handwritten signature recognition model for recognizing the handwritten signature based on the ConvNeXt module and the Vi T module; s4, performing first-stage fine tuning on a handwritten signature recognition model through the synthetic data set, and performing second-stage fine tuning on the handwritten signature recognition model through the real data set; and S5, identifying the handwritten signature through the fine-tuned handwritten signature identification model. The method has the advantages that the accuracy and robustness of handwritten signature recognition are greatly improved.

The invention discloses a federated learning method for tactical edge intelligence. A three-layer architecture of a data layer, a model layer and an application layer of tactical edge intelligence is formed. The method comprises the following steps: performing'federal 'type combination on computing nodes, dividing cluster partitions, strengthening a data layer to retain data locally, training and scheduling a local neural network model by a model layer aiming at multi-modal data of nodes of different partitions, performing customized aggregation on locally uploaded model parameters by an upper server, and performing multi-modal clustering on the locally uploaded model parameters by a lower server. Forming a distributed federal aggregation framework aiming at different partition differentiated data; the application layer provides a unified global model with high security and low communication cost for the upper part, and provides a customized personalized neural network model for edge node adaptation for the lower part; therefore, the communication frequency is greatly reduced, the communication efficiency is improved, and the problems of poor communication quality and network interruption are effectively avoided.

The invention relates to a method for realizing electronic screen writing personalization based on an AI writing model, and relates to the technical field of writing personalization, and the method comprises the following steps: obtaining writing data of a target user; standardizing the written data to obtain standard written data, and constructing a writing style model based on the standard written data; performing personalized analysis on the writing style of the target user by using the writing style model, and generating a personalized writing style; the personalized writing style is rendered to the target electronic screen through an electronic screen rendering module, and the personalized writing effect of the target user is displayed in real time. According to the method, the writing details of the user are analyzed through the deep learning model, the personalized writing style is generated, accurate reduction of the writing style is achieved, the personalized writing effect is rendered in real time in combination with the real-time input data and the style model, it is guaranteed that the fluency is consistent with the dynamic characteristics, and the writing experience and autonomy are improved.

Provided is a method (100) and system (200) for recent user-activity image based authentication. The system collects (202) recent user-activity and location data, generates (214) user-knowledge events from the recent user-activity according to location data, generates (421) random image questions concerning the user-knowledge events based on location. The random image questions test the user's knowledge of their recent user-activity according to location. A generative Applications Programming interface (API) (140) parses (220) the answered image questions, queries (222) the database for an expected response to the answered image questions, and compares (224) the expected response to the answers received to authenticate the user. In some embodiments a DALL-E (161) component and a GPT-4 V (162) component are used together in conjunction for such purposes. Other embodiments are disclosed.

A data processing system implements obtaining device information and performance requirements information for a resource-constrained computing device; analyzing the device information and the performance requirements information to determine an amount to compress one or more machine learning models to permit the resource-constrained computing device to operate the one or more machine learning models on the resource-constrained computing device, the one or more machine learning models including a stroke classification model for classifying digital ink stroke information as handwriting or a drawing; compressing the one or more machine learning models to permit the one or more machine learning models to operate on the resource-constrained computing device to generate one or more compressed machine learning models; and deploying the one or more compressed machine learning models to the resource-constrained computing device to process ink stroke information captured by a user interface of the resource-constrained computing device.

The invention provides a method for quickly acquiring an image splicing data set with high robustness. The method comprises the following steps: S1, adding characters on a blank image: adding characters with various gray levels on one blank image; s2, acquiring key points in the blank image: extracting the key points in the image by using an SIFT algorithm, and storing the positions of the key points for later use; s3, fusing various images with complex backgrounds with the blank background image: carrying out weighted fusion on the images with complex backgrounds and the blank background image in the step S2 by adopting a cv2. AddWeighted function in the OpenCV library to obtain a fused new image; and S4, generating a large number of image splicing training data sets with complex backgrounds and rich key point information through the image fused by the cv2. AddWeighted and the key point coordinate information formed by the SIFT algorithm in the step S2. Through the method, character image splicing training data with high complexity and high robustness can be generated in batches.

This disclosure relates to a hand-drawn graphic recognition method, apparatus, and system, and a computer-readable storage medium. The hand-drawn graphic recognition method includes: receiving an input of a hand-drawn graphic, the input including a stroke set, the stroke set including at least one stroke, each stroke including an information sequence of a plurality of trajectory points, information of each trajectory point including coordinates and writing time of the trajectory point; determining whether the stroke set forms a closed graphic according to coordinate data of the input stroke; and in the case where the stroke set forms a closed graphic, recognizing a category of the closed graphic based on the coordinate data of the stroke.

This disclosure describes systems, methods, and devices for analyzing and providing suggestions for handwritten math characters entered on a device A method may include receiving first handwritten strokes digitally entered on the device by a user; inputting the first handwritten strokes into a machine learning model; classifying, by the machine learning model, the first handwritten strokes as first mathematical characters; causing the device to present, based on the classifying, a selectable indication of the first handwritten strokes; receiving a first user selection of the selectable indication; causing the device to present, based on the first user selection, one or more selectable suggestions for second mathematical characters to follow the first mathematical characters; receiving a second user selection of a first suggestion of the one or more selectable suggestions; causing the device to present, based on the second user selection, synthesized second handwritten strokes representing the second mathematical characters.

A display apparatus displays handwritten data that is input with an inputter. The display apparatus includes processing circuitry; and a memory storing computer-executable instructions that cause the processing circuitry to convert the handwritten data into a character string; and conceal the character string generated by converting the handwritten data, in response to determining that a first time has elapsed after the character string is displayed.

Technology is described herein for facilitating a user's interaction with a digital ink document. The technology internally represents the ink document using a data structure having a hierarchy of nodes. The nodes describe respective elements in the ink document. The technology leverages the data structure to identify a set of nodes that grows upon the user's repeated selection of a particular part of the ink document. At each stage of the selection, the technology highlights a set of elements in the ink document that correspond to the current set of identified nodes. According to another illustrative aspect, the technology produces the data structure by modifying an original data structure provided by a text analysis engine. The technology performs this task with the objective of accommodating structured interaction by the user with the ink document.

A training method and device of an offline handwritten formula recognition model, equipment and a medium, comprising receiving a handwritten formula image containing a handwritten formula, determining a LaTeX label of the handwritten formula, and converting the handwritten formula into a printed formula image according to the LaTeX label; the printed formula image and the handwritten formula image are paired and input into a pre-constructed neural network model for training, the contrast learning loss of the trained neural network model is calculated according to the preset contrast learning loss function, and when it is determined that the contrast learning loss meets the preset condition, the trained neural network model is used as the offline handwritten formula recognition model, thereby using the paired input mode, and guiding the offline handwritten formula recognition model to learn semantic invariant features through the contrast learning loss function, thereby alleviating the influence of large writing style difference on performance and improving the overall recognition accuracy.

A method for generating a brush effect picture, an image editing method, a device, and a storage medium. The method includes: determining a pattern extraction area on a display interface in response to an area selection operation on the display interface of a display (S301); obtaining a first brush effect picture based on the pattern extraction area, where a pattern of the first brush effect picture is a pattern corresponding to the pattern extraction area on the display interface (S302); and obtaining a second brush effect picture based on the first brush effect picture, where the pattern of the first brush effect picture is the same as a pattern of the second brush effect picture, and definition of the second brush effect picture is greater than definition of the first brush effect picture (S303). A brush effect picture obtained by using the method has high definition. This improves user experience.

In some embodiments, an electronic device selects and provides for interaction with handwritten content in a content entry region. In some embodiments, an electronic device generates normalized shapes based on handwritten inputs. In some embodiments, an electronic device identifies actionable text within handwritten content and updates the identified text to be actionable. In some embodiments, an electronic device presents tutorials for text manipulation operations. In some embodiments, an electronic device displays visual feedback of text manipulation operations. In some embodiments, an electronic device selects units of content.

The application discloses a handwriting corner point determination method and device, a storage medium and a terminal, and relates to the technical field of computer image recognition. First, the writing speed of each trajectory point in handwriting and the final curvature of each trajectory point are calculated, and the final curvature is the average of all curvatures of vectors formed by the trajectory point and other trajectory points in a preset range. Then, a trajectory point meeting a condition is selected as a first corner point. Finally, pseudo corner points in the first corner point are removed. Since the average of all curvatures of the trajectory point in the preset range is taken as the final curvature of the trajectory point, that is, the average of the cumulative curvatures of the sliding window is taken as the curvature of the window center trajectory point, the misleading of curvature mutation to corner point selection is reduced, the curvature mutation can be effectively smoothed, more accurate corner points can be selected, line segment fitting of handwriting is performed, pseudo corner points in the first corner point are removed, the pseudo corner points in the corner points can be effectively removed, and finally, correct corner points in the handwriting are determined.

The invention discloses a handwritten electronic signature method and system based on multi-modal biological characteristics. The method comprises the following steps: carrying out real-name authentication on a signer before signing an electronic file; if the real-name authentication is passed, biological characteristic verification and evidence storage are carried out; recording handwriting data after biological feature verification is passed and signing is completed, and forming a data packet; the signed electronic file is printed into digital paper, the data packet is embedded into a digital paper file structure, and the digital paper is a layout file simulating real paper. According to the invention, the problem that the signature of the existing signer cannot represent the real identity and needs to depend on the CA certificate is solved, the tedious processes of applying, issuing, managing and the like for the CA certificate are avoided, and the online signature and the offline signature have the same efficacy.

Reference based digital content stylization techniques are described that enable generation of stylized digital content from a plain text input based on visual style attributes of reference digital content. A content processing system, for instance, receives input digital content including text and reference digital content with a particular visual style. The content processing system determines visual style attributes of the reference digital content, such as a text size, color, font, etc. The content processing system generates an input content classification scheme that includes semantic structure identifiers to classify text of the input digital content, e.g., as a header, subheader, paragraph, and so forth. Based on the input content classification scheme, the content processing system generates stylized digital content by applying the visual style attributes of the reference digital content to the text of the input digital content.