11 results about "Dynamic text" patented technology

This invention relates to the field of chromosome abnormality recognition technology, specifically to a chromosome abnormality detection method based on a multimodal large model. The method includes: constructing an image dataset and a text dataset; fusing image features and text features to obtain multimodal fusion features; assigning anomaly scores to image blocks belonging to band regions based on anomaly scoring rules formulated from the multimodal fusion features, and comprehensively processing the anomaly scores of all image blocks corresponding to the chromosome to determine whether the chromosome image is abnormal; and decoding and generating natural language text that meets the requirements of chromosome abnormality detection based on the multimodal fusion feature representation and anomaly scoring rules. This invention achieves accurate chromosome abnormality detection and band location positioning through multimodal fusion and dynamic text generation mechanisms, generating interpretable natural language text descriptions, and improving the practicality and interpretability of the detection results.

This invention discloses a text-guided and cross-attention-assisted method and system for re-identifying pedestrians changing clothes. The method includes: preparing a dataset for re-identifying pedestrians changing clothes; resampling using a diversity identity sampler, prioritizing samples with high diversity to construct training batches; building a cross-modal feature representation model based on contrastive language-image pre-training, generating dynamic text prompts, and performing visual and text cross-modal alignment training; extracting key biometric features using pedestrian parsing technology, and constructing a residual-cross-attention mechanism to fuse RGB global features with biometric features; performing a second-stage optimization training by combining identity loss, triplet loss, and adaptive salient feature loss; inputting a query image into the trained model to obtain a list of pedestrian targets ranked by similarity. This invention significantly improves the recognition accuracy and model generalization performance in solving problems such as uneven data distribution and lack of semantic information in clothing-changing scenarios.

User interfaces for indicating time, displaying a user interface based on a day of the week, user interfaces that include a dynamic text string, user interfaces that include a customizable border complication, user interfaces that include a watch hand that changes color at predetermined times, and user interfaces that include a simulated lighting visual effect.

This invention provides an open-vocabulary object detection method, apparatus, device, and storage medium, relating to the field of object detection technology. The method includes: extracting a visual feature map of the image to be detected based on an image encoder in an open-vocabulary object detection model; obtaining an alignment matrix of the visual feature map based on the positive dynamic text embedding, negative dynamic text embedding, and visual features of each target region in the visual feature map for each text category; wherein the positive and negative dynamic text embeddings for each text category are trained based on the visual features and easily confused visual features of that text category; and inputting the alignment matrix between each target region and each text category into a detection head to obtain the detection result. This invention improves the ability to distinguish easily confused categories during the detection process by establishing discriminative constraints for easily confused categories, thereby effectively improving the accuracy of open-vocabulary object detection results.

This invention relates to the field of document parsing, providing a method, apparatus, device, and storage medium for parsing tender documents. It utilizes a large language model to dynamically segment and extract semantic features from tender documents in a knowledge base, obtaining segmented text vectors. Based on the semantic features of the segmented text blocks, a multi-dimensional index structure is established according to multiple segmented text vectors. Pre-defined parsing items are converted into query commands through a question-and-answer interface of the knowledge base. Based on the query commands, the target information is obtained using the large language model and the multi-dimensional index structure. Compared to existing technologies where manual or simple tools struggle to effectively integrate the relationships within tender documents, this invention uses a large language model to extract semantic features from the tender documents and leverages the semantic association capabilities of the large language model to intelligently match and analyze the parsed tender requirements with the knowledge base, providing strong support for accurate recommendations and decision-making assistance.

This invention discloses a pedestrian re-identification method, medium, and device based on structured attribute-aware prompt learning, comprising: acquiring and preprocessing pedestrian image data, constructing a software-hardware collaborative dynamic text prompt template; inputting the pedestrian image and the corresponding software-hardware collaborative text prompt into a pre-trained image and text encoder to extract visual and text features, freezing the pre-trained network parameters, optimizing the learning of the software-hardware collaborative text prompt for each pedestrian, and inputting the corresponding frozen text encoder and trainable image encoder to obtain text and image embedding vectors; using the text embedding vector as a query and the image embedding vector as a key and value, performing depth alignment of visual and text features through N cross-modal networks to obtain the final fusion features; using the trained image encoder to extract pedestrian image features from the query image and image library images, and completing pedestrian re-identification retrieval by calculating feature similarity; this invention has high accuracy and strong generalization ability.

This invention discloses a zero-shot anomaly detection method based on Fine Interaction CLIP. A Fine Interaction CLIP model is constructed, in which a dual-path visual encoder includes a global path for extracting global visual features and a local path incorporating a relevance-aware attention mechanism for extracting local visual features. A dynamic text prompt construction module injects global visual features into learnable text prompts to generate text features. A feature refinement fusion module interacts with local visual features and text features to generate a deredundant prediction map, which is then fused with the global visual features to obtain a fused global feature. The model is trained using image-level and pixel-level loss. The trained model is then used to obtain image-level and pixel-level anomaly scores based on the fused global features and the prediction map. This method effectively addresses the problems of insufficient local visual feature extraction, weak static text prompt interaction, and false positives caused by feature redundancy, significantly improving the accuracy and localization precision of zero-shot anomaly detection.

The application provides a kind of automobile field entity recognition method and system based on large language model, the method extracts candidate entity in the text to be processed by first identification model, and the entity that cannot be mapped to standard knowledge base is marked as pending entity;When there is pending entity, retrieve relevant documents from the automobile field dynamic text library as supplementary knowledge, and conduct secondary entity recognition combined with large language model to output the final recognition result;The application adopts a hybrid knowledge base architecture, combines structured database and entity vector library to realize dual-channel retrieval, and supports dynamic incremental update based on the recognition result.The application effectively solves the problem of missed recognition caused by the strong timeliness and high ambiguity of professional terms in the automobile field, significantly improves the accuracy and recall rate of entity recognition, and at the same time, through the dynamic knowledge updating mechanism, the system's ability to quickly adapt to new car series and new brand release is guaranteed, which is suitable for application scenarios such as automobile vertical search and intelligent question answering.

The embodiment of the application provides a credit risk control method and equipment based on dynamic text information extraction, the method comprises the following steps: obtaining user information reflecting user credit, and determining the mandatory association level between the user information and the user and the dynamic influence level of the user information on the user according to the category of the user information; performing dynamic text recognition on the user information in the current stage to extract the first credit feature corresponding to the user in the current stage; performing risk assessment on the user to generate a lending strategy for the user according to the evaluation result of the risk assessment; selecting specified user information with a dynamic influence level higher than a preset influence threshold, and when the stage corresponding to the specified user information changes, performing dynamic text recognition on the specified user information in the changed stage according to the dynamic recognition mode to obtain the second credit feature corresponding to the changed stage; and modifying the lending strategy according to the compensated evaluation result.

This application relates to a typing competition method and system based on real-time ability modeling and dynamic text generation. The method includes: collecting user keystroke flow data in real time, calculating and weighting core and auxiliary ability indicators to generate a real-time ability vector; matching opponents based on vector similarity calculation and dynamic threshold; determining the basic text difficulty level based on both parties' abilities, and extracting the intersection of both parties' weak character sets as a common weak character set; generating original text based on this difficulty level, and replacing characters in non-critical semantic positions with common weak characters to form competition-specific text; synchronously transmitting the text to both parties' clients, dynamically adjusting the text difficulty by monitoring the input flow and exchanging cursor and error data in real time; after the competition, calculating a comprehensive score based on logs, generating a longitudinal report, and merging error character distribution data into the historical dataset to update the user's weak character set. This application improves the user's typing competition experience and training efficiency.

This invention discloses a semantically anchored, approximate contrastive learning method and system, relating to the fields of vision and language technology. It involves acquiring image and text data; extracting image embedding vectors and text embedding vectors using an image encoder and a text encoder, respectively; updating dynamic image and text queues through cross-modal momentum contrast and constructing a fused multimodal representation, which serves as a shared semantic anchor; progressively calibrating the image and text embedding vectors towards the shared semantic anchor within the contrastive learning framework to generate calibrated image and text embedding vectors; and performing image-text retrieval based on the calibrated image and text embedding vectors. This invention relaxes the strict requirement for complete semantic alignment of positive sample pairs, effectively mitigating cross-modal semantic bias.