23 results about "Smart memory" patented technology

The invention belongs to the field of natural language processing, and relates to an intelligent memory dynamic evolution method and system based on metadata and two channels, and the method comprises the steps: extracting an entity in an instruction input by a user, and obtaining an instruction entity; querying a metadata index database based on the instruction entity to obtain an entity state; when the entity state is a known state or a non-input state, updating the relational vector database and the graph database through a two-channel mechanism; a user input instruction is matched to the new relational vector database and the new graph database for mixed retrieval, and a preference relation and a hard constraint relation are obtained; constructing a context vector based on the user core portrait, the preference relationship and the hard constraint relationship; processing the user input instruction and the context vector to obtain generated content, and outputting feedback; and the interaction response capability and the engineering landing effect of the large language model intelligent agent in a long-period and complex interaction scene are greatly improved.

The invention relates to the technical field of display equipment, in particular to an intelligent lifting LED display screen support with a cable management function and a control method. The support comprises a base, and universal wheels with a locking function are arranged at the bottom of the base; the lifting stand column is composed of outer stand columns fixed to the base and inner stand columns capable of vertically moving relative to the outer stand columns, a connecting support is arranged on one side of each outer stand column, and the other end of each connecting support is fixed to the base. The fixed connecting plate is fixedly installed on the side faces of the top ends of the two connecting supports, a broadcast control box is arranged on the fixed connecting plate, and a power conversion circuit, a video sending card and a cable synchronous winding and unwinding mechanism are integrated in the broadcast control box; the screen body fixing frame is fixed to the top end of the inner stand column and used for installing an LED display screen body; and the intelligent driving control system is used for realizing lifting control, position memory and cable synchronous take-up and pay-off of the display screen. The problems that cables of an existing lifting support are exposed and disordered, adjustment is inconvenient, and an intelligent memory function is lacked are solved.

The application belongs to the field of natural language processing, and relates to an intelligent memory dynamic evolution method and system based on metadata and a double-channel, comprising the following steps: extracting entities in a user input instruction to obtain instruction entities; querying a metadata index library based on the instruction entities to obtain entity states; when the entity states are known states or unrecorded states, updating a relational vector database and a graph database through a double-channel mechanism; matching the user input instruction to the new relational vector database and the new graph database respectively for hybrid retrieval to obtain preference relations and hard constraint relations; constructing a context vector based on a user core image, the preference relations and the hard constraint relations; processing the user input instruction and the context vector to obtain generated content and output feedback; and greatly improving the interactive response capability and engineering landing effect of a large language model agent in a long cycle and a complex interactive scene.

The application relates to the technical field of artificial intelligence, in particular to an intelligent memory storage method based on multi-modal analysis and time sequence knowledge graph; the method is characterized in that multi-modal input information input by a user during interaction is textually analyzed to obtain a multi-modal text set; a text sequence formed by the multi-modal text set generated by multiple interactions is semantically disassembled to extract a knowledge unit set; a timestamp is added to the knowledge unit, the timestamp comprises an effective time interval of an event corresponding to an atomic fact extracted in the knowledge unit and a system recording time; a memory graph is constructed based on the knowledge unit with the timestamp, so as to search context information during subsequent interaction. The application has the effect of improving the accuracy of an intelligent agent in answering or helping a user to make a decision.

A reinforced concrete structure includes a first elongated concrete body extending in a first direction and a second elongated concrete body extending in a second direction, different from the first direction, and connected to the first elongated body. The first concrete body may include a first metallic reinforcement member embedded along its length. The second concrete body may include: a) plurality of second metallic reinforcement members, formed of a smart memory alloy, embedded along its length, and b) at least one third metallic reinforcement member, which may be formed of material other than a smart memory alloy, extending along its length. In a lateral cross-sectional area of the second concrete body, a cross-sectional area of the second and third metallic reinforcement members combined divided by the cross-sectional area of the third metallic reinforcement member ranges from about 3 to about 3.3.

The invention relates to the technical field of artificial intelligence. The invention provides a personal data intelligent memory management method, a personal data intelligent memory management system and electronic equipment. The method comprises the following steps: acquiring multi-modal original data of a user; performing preprocessing, macroscopic classification and label generation on the multi-modal original data to obtain structured multi-modal data, and constructing a knowledge graph by using the structured multi-modal data; performing intelligent management on the structured multi-modal data based on a three-layer memory management mechanism, and dynamically updating a knowledge graph; and calling a personalized inference model based on a query initiated by a user or a context perceived by a system, and extracting related information from the knowledge graph and the data after intelligent management to generate response content meeting user requirements. According to the method provided by the invention, deep mining and value utilization of personal data are realized by constructing the knowledge graph and introducing a three-layer memory management mechanism, the system can discover potential relations and laws among the data, and the user experience is effectively improved.

The invention belongs to the technical field of artificial intelligence, and particularly discloses an intelligent memory extraction device and method based on multi-dimensional association. The device comprises a memory unit structured storage module, a multi-dimensional association rule engine and a dynamic extraction and filtering module. The memory unit structured storage module is used for splitting memory information into independent memory units and storing the independent memory units in a vector database; the multi-dimensional association rule engine is used for defining and calculating association strength of the memory unit and the current theme in semantic, scene and emotion dimensions; and the dynamic extraction and filtering module is used for retrieving and screening the high-correlation memory unit to input the large language model according to the current theme. According to the method, on-demand extraction of the memory information is realized, the invalid context length is effectively reduced, and the model response efficiency and accuracy are improved.

An artificial intelligence (AI) memory processing method includes acquiring multimodal data; and converting the multimodal data into parametric memory data and non-parametric memory data, where the non-parametric memory data includes memory data stored in a memory unit, and the parametric memory data includes memory data stored in a parameter of an AI model. An AI memory processing device includes at least one processor and a memory unit communicatively connected to the at least one processor; where the memory unit stores a computer program executable by the at least one processor to enable the at least one processor to perform the preceding method.

A smart memory module includes at least one memory device connected to at least one channel, a controller configured to control the at least one memory device, and a host interface circuit configured to perform peer-to-peer (P2P) communication with an external device, and to receive, from a host device, a live migration request. The controller is configured to perform, in response to the live migration request, a migration operation of transmitting buffer data to a target device through the P2P communication under a buffer-full condition.

The invention discloses an intelligent memory and target identification method based on micro-model self-learning, and relates to the technical field of intelligent and computer cognitive computing. Comprising the steps of initial target collection and feature extraction, constant-level micro-model memory construction, multi-model collaborative recognition and memory calling, micro-model autonomous updating and growth and memory management and recognition result output. According to the method, the intelligent exclusive memory device is constructed, and the tracking model and the micro-model self-learning ability are fused, so that the technical effects of constant-level space storage-target feature autonomous updating-recognition model growth while using are realized. According to the method, a certain micro-model scale is fixed, the traditional linear storage bottleneck is broken through, meanwhile, the continuous learning ability is achieved, and the storage and adaptation problems of the intelligent equipment in target memory and recognition are solved.

The utility model provides a transmission device, and especially relates to an OTG intelligent memory storage transmission charging line comprising a first transmission assembly, a second transmission assembly, a connection assembly and a storage device. The connecting assembly is connected with the first transmission assembly and the second transmission assembly. The storage device is arranged in the connecting assembly, and when the first transmission assembly is connected with the first equipment and the second transmission assembly is connected with the second equipment, the second equipment and the storage device are in communication connection with the first equipment through the first transmission assembly at the same time.

The invention provides a vehicle-mounted terminal split-screen display method, a vehicle-mounted terminal split-screen display device, a vehicle and a medium, and the method comprises the steps: responding to a split-screen request of a user for a target split-screen application, and judging whether the split-screen state cache of the target split-screen application is valid or not according to a split-screen state detection condition; wherein the target split-screen application refers to an application which has entered a split-screen mode within historical time, and the split-screen state detection condition comprises an application self-hold condition and a split-screen non-interruption condition; and if the target split-screen application is valid, performing split-screen display on the target split-screen application in a display window according to the split-screen layout configuration in the split-screen state cache. By means of the method and device, the intelligent memory and one-key restoration function of the split screen state is achieved, the tedious operation that a user repeatedly enters the split screen mode and repeatedly selects applications is avoided, the interaction steps of the user are remarkably reduced, and the multi-task switching efficiency is improved.

The application relates to a shape memory sponge and a preparation method and application thereof, the shape memory sponge comprising a polyurethane memory sponge and a modified nanocomposite, the modified nanocomposite comprising modified nanometer copper particles and titanium oxide nanoparticles; wherein the modified nanometer copper particles have a core-shell structure, the core is copper nanoparticles, and the shell layer is an antioxidation coordination shell layer coated outside the copper nanoparticles. The scheme provided in the application has the intelligent memory effect and long-acting broad-spectrum antibacterial and antiviral performance.

The application relates to the technical field of artificial intelligence. A personal data intelligent memory management method, system and electronic equipment are provided. The method comprises the following steps: acquiring multi-modal original data of a user; performing preprocessing, macro-classification and label generation on the multi-modal original data to obtain structured multi-modal data, and constructing a knowledge graph by using the structured multi-modal data; intelligently managing the structured multi-modal data based on a three-layer memory management mechanism, and dynamically updating the knowledge graph; based on a query initiated by the user or a context perceived by the system, calling a personalized reasoning model, extracting relevant information from the knowledge graph and the intelligently managed data to generate response content meeting the user demand. The method provided by the application realizes deep mining and value utilization of personal data by constructing a knowledge graph and introducing a three-layer memory management mechanism, the system can discover potential connections and rules between data, and the user experience is effectively improved.

This invention discloses a dead reckoning system and method for a dual-wheel differential AGV, belonging to the field of dead reckoning technology. The system employs a data acquisition module to acquire multi-source operating data of the AGV and preprocess it to generate a standardized data stream; a mileage calculation module to generate the original mileage increments for the left and right wheels; a heading fusion module to fuse the standardized data stream and the original mileage increments through a filter to generate a fused heading angle; a motion control module to generate motor drive commands based on the fused heading angle and target commands; a memory optimization module to identify scenarios and match historical strategies, dynamically adjusting relevant calculation parameters; and a visualization module to visualize the operating data and control process and generate a trajectory report. This invention, by fusing multi-source sensor data and intelligent memory optimization, achieves adaptive parameter adjustment and scenario-based strategy reuse in dead reckoning, effectively suppressing accumulated errors and improving the motion control accuracy and robustness of the AGV in complex environments.

A method for customized smart memory recap for viewed content is disclosed. The method includes: receiving input content into the system through integration with a content delivery system; processing, via a content recognition engine, one or more of metadata, captions, and video / audio feeds associated with the input content to determine significant scenes; extracting, via the content recognition engine, data from the input content associated with the significant scenes; identifying and tagging, via a key moments extraction module, key moments in the significant scenes in the input content; compiling, via a summary generation engine, a summary recap using tagged key moments associated with the significant scenes; customizing, via the personalization module, the summary recap based on user data collected from user input while viewing the input content, and outputting, via a delivery system, the customized summary recap to the user.

This invention provides a method for assessing embryonic development quality based on a Smart Memory Network (SMMC). The method involves inputting embryonic images at different developmental stages; assigning different weights to the embryonic images according to the developmental stage, and outputting an attention feature map with channel attention; extracting global and local features from the attention feature map; fusing the global and local features and extracting temporal features to obtain a spatiotemporal feature map; and inputting the spatiotemporal feature map into a binary classification fully connected layer for classification to assess embryonic development quality. By extracting features from various aspects, the method of this invention exhibits superior performance in embryo assessment.

1. Name of the product in this design: EEG Intelligent Memory Eye Mask (Stardust Texture). 2. Purpose of this design: It is mainly used as an eye mask for use in scenarios such as EEG signal acquisition, sleep monitoring, meditation and relaxation, and intelligent sleep aid. 3. The key design feature of this product is its shape. 4. The image or photograph that best illustrates the design's key points: a 3D model.

Provided are a boot controller and a computing device including the boot controller. After receiving a power-on-reset signal from the computing device, the boot controller verifies the integrity of boot code stored in read only memory (ROM) of the computing device by reading the boot code from the ROM, and determines a boot mode of the computing device based on results of the verification of the integrity of the boot code stored in the ROM.

This invention discloses a dynamic prism memory-based scene generation method in the fields of computer graphics and virtual reality, addressing the problems of existing scene generation technologies, such as lack of effective memory mechanisms and dynamic adaptability, high time cost, large memory consumption, and poor real-time performance in generating complex scenes. It employs an octahedral geometric structure as a logical storage model, mapping eight types of scene data, including geometric, texture, and lighting information, to eight logical storage areas, maintaining data consistency through an 8×8 symmetric association matrix. An intelligent memory navigation matrix is ​​established, using a locality-sensitive hashing algorithm combined with a B+ tree index structure for scene matching. Progressive scene memory fusion is performed, calculating scene differences through a scene differential encoding algorithm and integrating information using an adaptive weight adjustment factor. A hierarchical update strategy is adopted to output the generated scene based on the degree of scene change. Scene generation speed is improved by 52%, and memory consumption is reduced by 26%.

A mechanism is described for facilitating memory handling and data management in machine learning at autonomous machines. A method of embodiments, as described herein, includes detecting multiple tables associated with multiple neural networks at multiple autonomous machines, where each of the multiple tables include an index. The method may further include combining the multiple tables and multiple indexes associated with the multiple tables into a single table and a single index, respectively, where the single table is communicated to the multiple autonomous machines to allow simultaneous processing of one or more portions of the single table using one or more memory devices and one or more processors of one or more of the multiple autonomous machines.