7results about How to "Prevent error propagation" patented technology

The application discloses a kind of man-machine collaborative multi-modal data intelligent labeling and quality checking method, it is related to man-machine collaborative technical field.The method first obtains at least containing two modalities of original data of image, text, audio, generates multi-modal joint data sample after pre-processing and modal alignment;Input pre-trained multi-modal AI pre-labeling model generates pre-labeling result with confidence;Through cross-modal consistency check calculation consistency score, sample is divided into high, medium, low confidence three categories;High confidence sample is directly adopted and enters the sampling pool, and medium and low confidence sample is based on labeling personnel ability portrait and AI model dynamic capability evaluation and carries out intelligent task distribution.The application effectively solves the problem that efficiency and quality are difficult to balance in multi-modal data labeling, significantly improves labeling accuracy and automation rate, reduces labeling cost, and can meet the construction demand of large-scale high-quality multi-modal data set.

This invention discloses a method for extracting parameters from deep neural networks under constrained architectural knowledge conditions. Its key feature is that before extracting parameters from the current hidden layer, it analyzes the algebraic consistency between the perturbation basis matrix and the output difference vector in the input space to accurately identify whether there are any missing neurons in the previous layer. When a missing neuron is detected, a re-extraction mechanism is triggered, increasing the sampling density of key points and re-performing parameter recovery of the missing layer. Compared with existing technologies, this invention achieves efficient and accurate extraction of DNN parameters under constrained architectural knowledge conditions through an adaptive extraction workflow and neuron omission detection. Its error detection recall rate reaches 100%, and its error detection specificity remains stable between 1.91% and 2.03%. This provides a robust and efficient technical solution for model vulnerability assessment, black-box model parameter extraction attack prevention, and neural network intellectual property protection in the field of artificial intelligence security, and has broad application prospects and commercial value.

The application discloses a kind of human in loop multi-agent microservice system AgentInn design and construction method, including human in loop control module, multi-agent cooperation module, dynamic capability expansion module and task arrangement module.Human in loop control module realizes the control, participation and observation of human to agent by starting confirmation, running intervention, abnormal recovery, human agent registration, cooperation routing and experience feedback;Multi-agent cooperation module realizes cooperation arrangement and load balancing by cooperation space, role load balancing and agent discovery;Dynamic capability expansion module supports on-demand creation and hierarchical exposure of tool service and intelligent service;Task arrangement module manages progress using Cycle-Sprint-Phase three-layer time abstraction, and realizes task flow by combining flow decision and context transmission.The application solves the problems of lack of human participation in multi-agent cooperation, high complexity of cooperation arrangement and limited capability expansion, and realizes multi-agent microservice cooperation.

This invention discloses a method for managing encrypted write anomalies in enterprise-level storage systems. During the data write process, based on the security requirements of the data and the operational status of the storage unit, a corresponding encryption strategy is dynamically selected. During the encrypted write process, reversible encryption markers or failure markers are established for data blocks. When a data block is written but has not reached stable write conditions, its encryption and write phase states are recorded using reversible encryption markers. In subsequent management operations, the process is rolled back to the previous stable point, reconfirming the correspondence between the encryption strategy and the key version without changing the original encrypted content. When a data block write fails, re-encryption and cross-storage unit compensation write processing are triggered using failure markers. By performing distinct management operations based on different markers, this invention can effectively reduce the impact of write anomalies on system stability and performance while ensuring data security.

The present application relates to the technical field of action control, in particular to a continuous action control method and system based on fork loading and unloading, the present application realizes adaptive identification of cargo unloading event by collecting torque feedback of lifting driving mechanism and performing differential analysis, and latches actual unloading fork position; based on the position, middle position and target loading position, a complete motion path is constructed, and two-stage stroke compatibility determination of fork retraction and extension is performed to determine whether continuous action is enabled; during continuous action, fork extension and walking shaft motion are scheduled in parallel, and subsequent fork extension action is triggered through state flag synchronization mechanism; at the same time, motion parameter is used to predict loading completion time and send pre-notification message; in abnormal conditions, a cancellation message is sent and a safety rollback is performed. The present application can effectively improve the continuity, efficiency and safety of fork operation in automatic warehouse system.

The application discloses a multi-agent task cooperative processing system and method for voice interaction. The system comprises: a voice input module for receiving and preprocessing original audio signals from a user; a voice processing module for generating a structured task representation based on the audio signals using a deep neural network model; a task coordination module for assigning tasks to agents in a task processing module based on the structured task representation; the task processing module contains multiple agents, all connected to the task coordination module, each configured to process tasks in a specific field; a response generation module for integrating the processing results from the agents into a coherent response and converting it into a voice signal for output. The application integrates voice understanding, task scheduling and cooperative execution to provide a robust, efficient and intelligent voice-driven solution that can complete complex voice interaction tasks.

This application provides a method, device, and electronic device for managing advertising campaign costs based on a dual-track online and offline classification system. By integrating a dynamic knowledge graph with a prompting learning model that efficiently fine-tunes parameters, it achieves high-precision automatic classification of advertising campaign cost categories. The method acquires the user-inputted cost description text and specified category, utilizes the dynamic knowledge graph for key entity extraction and multi-hop reasoning, and generates a first classification result and confidence score. Simultaneously, the text is input into the prompting learning model, which guides fill-in-the-blank prediction through soft prompts, outputting a second classification result and confidence score. Combining the confidence scores of both, a decision fusion rule is used to determine whether to trigger user correction prompts, and the knowledge graph and model parameters are dynamically updated based on user feedback. The system supports new entity recognition and incremental learning, forming a closed-loop optimization mechanism. This solution effectively solves the problems of inaccurate budget allocation, low execution efficiency, and difficulty in cost control caused by the fragmentation of data from online and offline channels, improving the accuracy, interpretability, and adaptability of cost classification, and achieving intelligent and refined advertising campaign cost management.