4results about How to "Easy to audit" patented technology

The invention relates to the technical field of computers, and discloses an intelligent behavior data analysis method based on artificial intelligence. Collecting data of at least two data sources, generating an original record identifier for the original behavior record, and associatively storing the original record identifier as an original evidence record set; preprocessing the record set to obtain a behavior event sequence, segmenting the behavior event sequence into behavior segments, and constructing a behavior primitive library; and extracting semantic features and time sequence features from the fragments, determining a candidate primitive set, inputting the time sequence features into a first artificial intelligence model to determine a target primitive, and fitting parameters to obtain a primitive mapping result. And performing consistency verification based on the constraint set, and sequentially replacing the target primitive and adjusting the boundary of the fragment to perform fallback recalculation when the verification is not passed. And generating an evidence tuple for a verification passing result, linking the evidence tuple into an evidence packet, recording a feature snapshot and a version identifier, inputting a primitive time sequence chain into a second artificial intelligence model to output a behavior analysis result, and performing recomputable playback based on the evidence packet to reproduce the analysis result.

This application provides a target re-identification method, apparatus, and medium, relating to the field of computer technology. The method includes: acquiring first multimodal features of candidate target pairs, the first multimodal features including physical color features, appearance vectorization features, specified structured features, and spatiotemporal context features; the specified structured features including specified structured information and its recognition credibility; and the spatiotemporal context features including the physical velocity of the candidate target pair; inputting the first multimodal features into an improved XGBoost model to determine whether the candidate target pair is the same target, including directly determining whether the candidate target pair is the same target based on the specified structured information whose recognition credibility is greater than a preset threshold and the physical velocity of the candidate target pair satisfying a preset short-circuit rule. This application extracts multimodal features of candidate target pairs in the first stage, uses gradient boosting trees for target re-identification in the second stage, and sets a short-circuit rule mechanism to save computational power and improve system robustness and interpretability.

This application discloses a data-aware intelligent analysis method, system, storage medium, and computer device. The method includes: receiving a target problem and initializing state information, wherein the state information includes key sub-conclusions, a reasoning state logic diagram, and key data descriptions; a language model-based agent iteratively performs the following operations: obtaining input information for the current step; updating the state information of the current step based on the action and environmental feedback information of the previous step, and making decisions based on the state information of the current step to obtain the action and tools to be invoked for the current step; invoking the tools to execute the action of the current step, and obtaining environmental feedback information for the current step after the action is completed; checking whether the target problem has been analyzed; if not, proceeding to the next loop; if completed, generating analysis results. This method can improve the reasoning ability, response consistency, and interpretability of deep reasoning models.

This invention discloses an individual motion behavior analysis system based on dynamic modal configuration and rule-guided fusion, belonging to the fields of intelligent perception, sports science, and human-computer interaction. It includes a data input module, a dynamic modal organization and configuration module, a multimodal feature modeling and fusion module, an indicator-level evaluation and behavior state prediction module, a statistical rule-guided analysis and result correction module, and a structured result generation and output module. These modules work together to form a complete closed loop for motion behavior analysis technology. This invention aims to address the problems in existing motion behavior analysis technologies, such as limited dimensionality of single-modal information, rigid multimodal fusion methods, insufficient stability and interpretability of analysis results, and strong dependence on large-scale manually labeled data.