27 results about "Visual cognition" patented technology

The invention discloses a fabric defect visual detection method based on fabric visual cognition, and relates to the technical field of textile production and quality control, and the method comprises the following steps: generating standardized fabric surface image data; constructing texture structure features used for representing normal fabric texture distribution; generating a fabric defect candidate area; obtaining a fabric defect detection result; and constructing a time sequence feature analysis model, generating corresponding quality risk early warning information and prevention and control suggestions, and generating batch-level quality inspection associated data. According to the method, the problems that in the prior art, fabric defect feature representation is not comprehensive, the complex defect recognition accuracy is low, and quality inspection data are difficult to continuously store and trace are solved, the fabric visual cognition model based on fabric visual cognition is constructed, and multi-dimensional texture structure features and a time sequence analysis mechanism are fused; according to the invention, accurate identification and quality risk prediction of fabric defects are realized, and the technical effects of stability, continuity and traceability of fabric quality detection are improved.

Disclosed are examples of training-free systems, methods and apparatuses, rooted in Chainof-Thought (CoT) reasoning, used to enhance the zero-shot performance of vision language models (VLMs) such as CLIP on a variety of downstream tasks. Hierarchical questions reflecting human visual cognition can be used with a pre-trained visual question answering model to extract the context of a query image from a global to local perspective through strategic questioning. Those CoT-based question-answer (QA) pairs, in conjunction with predefined class names, can serve as input to a language encoder, resulting in multi-level textual embeddings that emphasize various aspects of the image to improve existing VLM performance without additional training or labelled data.

The invention relates to the technical field of depth image clustering, in particular to an interpretable depth image clustering method based on dynamic feature routing. The method comprises the following steps: acquiring an input image, and extracting visual features of the image by using a pre-training model; the visual features and the text features are processed to generate a similarity graph, and whether the model serves as a backbone to learn the visual features consistent with human visual cognition or not is verified; inputting the visual features into the explainable feature router, and aligning the learnable weight of the explainable feature router with the visual features through a dynamic feature routing mechanism; the dynamic feature routing mechanism scales the weight according to the similarity between the weight vector and the input feature vector; based on the aligned features, feature extraction and clustering distribution are carried out through a feature head and a clustering head; and performing joint optimization on the interpretable feature router, the feature head and the clustering head based on a manifold linearization and clustering joint objective function. According to the invention, the accuracy of the depth image clustering result is improved.

The utility model relates to the technical field of educational toys for children, and discloses a cognitive educational toy for children, which comprises a shell, a groove is arranged in the shell, a drawer is slidably connected in the shell, a partition plate is fixedly connected in the drawer, jigsaw blocks are arranged in the drawer, splicing components are arranged on the side walls of the jigsaw blocks, and the splicing components are arranged on the side walls of the jigsaw blocks. A fixing assembly is arranged in the drawer; each splicing assembly comprises a first clamping block, the first clamping blocks are fixedly connected to the side walls of the jigsaw blocks, clamping grooves are formed in the jigsaw blocks, and first magnetic attraction blocks are fixedly connected to the interiors of the jigsaw blocks. According to the jigsaw puzzle, the grooves of different shapes are formed in the shell, the first clamping blocks slide into the clamping grooves of other jigsaw puzzle blocks, the first magnetic attraction blocks and the second magnetic attraction blocks are matched for fixing, and then patterns of different shapes corresponding to the shapes of the grooves are formed, so that the visual cognition of children on different shapes is improved, and the interestingness of children is improved. Through the structure, the spatial imagination and spatial reasoning ability of children can be cultivated.

The invention relates to the technical field of light atmosphere, and discloses a light atmosphere effect generation method, which comprises the following steps: generating an initial wave function according to an intensity field, deducing on the basis of a potential energy field to obtain a first wave function, performing intensity calculation and phase extraction from the first wave function to obtain a second wave function, updating the direction field based on the second wave function to obtain an updated direction field and an enhanced optical flow field; three core advantages of light atmosphere generation are realized, firstly, deep coupling of quantum optics and neurocognition is realized, real response of human eyes is simulated through a phase-sensitive light response model, and light atmosphere distribution is enabled to conform to quantum propagation rules and human visual cognition characteristics in combination with a cognition attention focusing mechanism of a DQN model; and 2, a dynamic self-adaptive mechanism runs through the whole process, and a light current coefficient and object motion data are adjusted in real time through the pupil diameter to feed back a new light flow field so as to ensure that the environment change can be quickly responded.

The invention relates to a visual thinking chain, in particular to a bionic visual thinking chain cross-modal model based on eye movement tracking and a visualization system and method, and solves the technical problems that an existing visual thinking chain model lacks real visual thinking process simulation and is difficult to establish a visual thinking chain conforming to a real human visual thinking mode. According to the bionic visual thinking chain cross-modal model based on eye movement tracking, the to-be-detected image eye movement gazing area image blocks from eye movement tracking data and the gazing sequence and the gazing duration of the to-be-detected image eye movement gazing area image blocks are adopted, and the dynamic characteristics and the time sequence mode of human visual cognition can be truly reflected; the big language model is made to process visual information step by step according to the sequence of human eye movement tracks, a human real visual thinking mode is simulated, and therefore a real visual thinking chain is established; besides, an eye movement vector prediction matrix and an eye movement prediction text answer can be obtained, so that a bionic visual thinking chain and a text thinking chain are output at the same time, and the interpretability of the thinking process after the text and the image are fused is improved.

The invention discloses an interactive cognitive impairment detection device, which is applied to the technical field of postoperative cognitive impairment detection.The interactive cognitive impairment detection device is provided with a display controller, virtual reality interactive glasses, a tactile feedback device and an interactive cognitive impairment detection system, so that a medical worker can perform interactive cognitive impairment detection work with a patient; visual, listening, touching, sniffing and other feelings can be added, the situation that the cognitive function of the patient is evaluated only through languages, charts, characters and the like is avoided, therefore, comprehensive judgment can be conducted through added visual cognition, touch, hearing and even smell, the cognitive level of the patient can be judged more accurately, the limitation of the cultural level of the patient on the test is avoided as much as possible, and the test efficiency is improved. Moreover, interaction with the patient can be added, so that more accurate evaluation is realized, and the occurrence proportion and degree of the postoperative cognitive impairment of the patient can be evaluated more accurately and scientifically.

Provided are: a nucleic acid including a nucleic acid sequence encoding a chimeric protein including at least part of an ion-transporting receptor rhodopsin and at least part of a G protein-coupled receptor rhodopsin and a nucleic acid sequence encoding a signal sequence; and a nucleic acid including a nucleic acid sequence encoding a chimeric protein including at least part of an ion channeling receptor rhodopsin and at least part of a G protein-coupled receptor rhodopsin; and a nucleic acid construct including the nucleic acid sequences. The use of the nucleic acids or nucleic acid constructs prevents and suppresses the progress of retinal diseases, and enhances the visual cognitive behavioral function and visual function.

The application discloses a fine-grained visual target recognition expert knowledge intelligent agent generation method and device, and the method comprises the following steps: generating a first type of prompt word of a to-be-recognized image based on feature sorting, feature weight, and semantic description information of a diagnostic feature in the to-be-recognized image, wherein the feature sorting and the feature weight are determined according to subjective data and objective data of experts in identifying the same target; inputting a field to which the to-be-recognized image belongs, the to-be-recognized image and the first type of prompt word into a cognitive process reasoning system based on a large language model to obtain reasoning text, wherein the reasoning text is used for simulating a visual cognitive process of an expert in identifying the to-be-recognized image. The application can utilize existing part of expert field knowledge to drive a large model to efficiently and automatically generate reasoning text simulating a human expert identification fine-grained target process.

The application provides a game auxiliary display method and display screen based on visual cognitive state, which comprises the following steps: collecting eye movement signals of eyeballs of players in real time and extracting visual feature parameters therefrom; calculating an intensity index of visual tunnel effect and dividing visual cognitive state grades; identifying effective information in a game picture, screening out key information according to the influence weight of the effective information on the game, and marking spatial coordinate parameters of each key information; mapping the key information to corresponding microlens display areas of the screen; and driving the microlens display areas to output auxiliary display pictures. The scheme constructs a dynamic visual cognitive state of players and a self-adaptive matching system of directional auxiliary display of the screen, accurately matches dynamic change rules of the visual tunnel effect under high-intensity competitive games, improves the effective capturing efficiency of human eyes on key early warning information of the game and the information transmission timeliness under the premise of completely retaining the original picture visual effect of the game and not interfering with the core operation rhythm.

The present application relates to the technical field of medical health informatics, in particular to a myopia early warning method and device, a storage medium and an electronic device; the method comprises the following steps: through wearing a device integrating a thermal sensor and a nerve conduction film on the face of a user, static acquisition of eye thermal distribution gradient and nerve response time delay data is used to construct an individualized baseline model; dynamic monitoring of thermal field deviation, nerve conduction change and electromyographic disturbance in visual cognitive tasks; combination of a dynamic deviation mapping algorithm and an adaptive weight distribution model to calculate a multi-dimensional risk index, prediction of a myopia-induced fatigue risk level, and matching of a personalized intervention strategy according to user portrait characteristics, to realize closed-loop early warning and intervention optimization. The present application synchronously analyzes physiological signals and behavior data through real-time acquisition, and executes data processing and early warning decision-making in combination with an electronic device and a storage medium, thereby effectively improving the accuracy of early myopia warning and the intervention adaptation degree.

This invention discloses a multi-scale feature fusion method for defect recognition in injection molded parts, belonging to the field of computer vision and industrial product quality inspection technology. The method includes: constructing and preprocessing a dataset of defective injection molded parts images; extracting multi-scale feature maps using a deep convolutional neural network with an integrated visual cognitive computing module; enhancing small-target defect information by collaboratively modeling the consistency and differences of features through a parallel dual-branch feature fusion module; employing an attention-guided cross-level enhancement module to reweight the fused features through channels, highlighting key features; and finally, achieving accurate classification and localization of defects using a detection head network based on adaptive anchor boxes. This method can effectively improve the accuracy and recall rate of injection molded parts, especially for minute defects, and is suitable for online quality inspection on industrial production lines.

The invention relates to the technical field of eye movement data processing, in particular to a cognitive disorder assessment and recognition method and system based on eye movement data. The method comprises the following steps: acquiring eye movement behavior data of a plurality of fixation points when a user executes a visual task; dividing the duration of the visual task into a plurality of segments, and constructing a visual-cognitive coupling factor of each segment; calculating a mean value of the coupling factors in each segment; performing linear regression on the mean value, and taking the slope of a regression straight line as a dynamic adaptability index; and forming a feature vector by the dynamic adaptability index and the eye movement behavior data, and inputting the feature vector into a trained classification model to obtain a risk cognitive state of the user. According to the scheme, the accuracy of cognitive impairment evaluation and recognition can be improved.

The application discloses an emergency evacuation design and rescue method for large-scale comprehensive buildings, and particularly relates to the field of emergency evacuation and rescue design, and is used for solving the problem that the rescue path planning and people flow guidance are disconnected due to the complex internal space structure of the existing comprehensive buildings and the lack of visual cognitive support in traditional evacuation design; high-frequency stay areas are identified by collecting the crowd moving tracks of multiple comprehensive buildings, consensus space memory features are extracted and a space memory anchor network is constructed, evacuation path nodes are labeled in the network and weak visual connection sections are identified, new memory anchor points are formed by arranging micro fire stations, the guide utility coefficients are calculated in combination with fire drill data and are converted into dynamic conflict weights, when a fire event occurs, rescue path search is performed in the anchor network with the target of minimizing the cumulative dynamic conflict weights, and the evacuation guidance optimization based on visual memory features and the dynamic cooperative design of the rescue path are realized.

The application belongs to the technical field of digital twin cities, and discloses a method and system for quantitatively evaluating the visual cognitive chaos of a user in a digital twin scene, which comprises the following steps: acquiring eye movement sequence data of a target subject in the digital twin scene and pre-processing the eye movement sequence data; determining a reference frame of the eye movement sequence data, reconstructing the phase space of the pre-processed eye movement sequence data based on the reference frame; calculating the nearest neighbor points of all observation points in the eye movement sequence data in the phase space, and calculating the average logarithmic separation curve through the observation points and the nearest neighbor points; obtaining the fitting slope of the average logarithmic separation curve based on a linear regression method, determining the chaos degree of the target subject based on the fitting slope; and outputting the chaos degree determination result of the target subject; the application can quantitatively detect the visual cognitive chaos of the target subject in the digital twin scene, and can be applied to the fields of traffic safety monitoring, pedestrian risk prediction and human-computer interaction optimization.

This invention relates to the fields of network security and big data analytics, specifically to a network security information big data analytics system capable of comparing historical data. The system includes: a historical baseline extraction module, used to acquire real-time security data containing the current data node and corresponding long-term historical security baseline data, generating the anomaly weight and risk density of the current data node; a visual cognitive assessment module, communicatively connected to the historical baseline extraction module, generating adaptive visual adjustment instructions; a cognitively driven adaptive rendering module, communicatively connected to the visual cognitive assessment module, generating a dynamic security map after cognitive denoising; and a spatiotemporal evolution dynamic rendering module, communicatively connected to the cognitively driven adaptive rendering module, based on a three-dimensional evolution holographic view presenting long-term attack paths. This invention avoids visual overload for analysts caused by massive concurrent data volumes.

The invention relates to the technical field of forest damage detection and image processing, in particular to a system and a method for automatically detecting plant diseases and insect pests of a needle-leaved pure forest by fusing multi-modal visual cognition modeling and unmanned aerial vehicle remote sensing, which comprises the following steps: collecting canopy images and metadata through unmanned aerial vehicle visible light remote sensing, and innovatively constructing a pathological feature enhancement model; extracting a non-green area mask through HSV color gamut threshold segmentation, and enhancing pathological features in combination with contrast gain and a brightness suppression coefficient; then, a vision-language collaborative reasoning framework is constructed, enhanced vision information and pathological description cues are fused, and zero sample recognition of diseases and insect pests is achieved; and finally, a detection result is efficiently output through an asynchronous processing pipeline, and the regional damage rate is calculated. According to the method, pathological feature enhancement and multi-mode reasoning are integrated, the detection precision, the positioning precision, the efficiency and the damage rate estimation error of the method are all superior to those of a conventional model in the needle-leaved pure forest environment damaged by the pine bark beetles, high efficiency and reliability are verified, and a new way is provided for monitoring the diseases and pests of the needle-leaved forest.

The invention relates to the technical field of image evaluation indexes, in particular to a context-based camouflage target segmentation measurement method, which comprises the following steps of: defining pixel correlation intensity distribution based on a Gaussian law according to a manual annotation graph; carrying out forward reasoning (calculating the understanding degree of a model prediction result on a real target result through convolution approximation) and reverse deduction (verifying the reflection of the real target result in the model prediction result through convolution approximation and normalization processing), and carrying out weighted score calculation on the results of forward reasoning and reverse deduction; and finally, an evaluation result of the disguise target segmentation is obtained, the evaluation result keeps higher consistency with human visual cognition, and the actual performance of the image segmentation model in the disguise scene can be measured more comprehensively and accurately.

The invention relates to an automobile design benchmarking picture retrieval method and system based on a three-dimensional classification framework, and belongs to the field of picture retrieval. The method comprises the steps that an automobile design benchmarking picture library is acquired, and a probability generation type three-dimensional classification tree is constructed through dynamic tree splitting; performing semantic-visual cognition manifold alignment based on a probability generation type three-dimensional classification tree, and outputting a joint cognition space and double encoders; the joint cognitive space and the double encoders are optimized through user feedback; based on the joint cognitive space and the double encoders, automobile design benchmarking picture retrieval is carried out in the automobile design benchmarking picture library, multi-dimensional attributes are fused, design evolution is dynamically adapted, semantic and visual alignment is achieved, continuous optimization can be achieved through user feedback, and the efficiency and quality of design benchmarking work are improved.

The invention discloses a cognitive driving space behavior modeling method based on inverse reinforcement learning. The method comprises the following steps of space behavior data collection, initial agent setting, environment cognitive feature extraction, space behavior modeling, behavior mechanism analysis and data library building and visualization. According to the method, a space-behavior model capable of reflecting visual cognition, functional cognition and community cognition is formed by introducing a video track and space environment data and combining optimal path baseline setting and inverse reinforcement learning training; accurate prediction and mechanical explanation of crowd behaviors in a public space can be realized, and the method is suitable for application scenes such as space design optimization, group behavior simulation and smart city management.

The invention relates to a visual cognition-driven real-time intelligent electronic note arrangement method, which adopts two stages of processes, in the first stage, a Transform-based layout generation model with preferential readability is adopted, and in the first stage, the length-width ratio and the size of an original screenshot are mainly reserved, and an initial readable layout is generated; in the second stage, cognitive-driven readable layout adjustment is carried out, and a layout logic sequence is optimized through a tree-based sequence adjustment strategy so as to be matched with human reading habits. According to the electronic note collage layout generation method and device, the recognition science principle and the deep learning technology are deeply fused, the problem of contradiction between readability and cognition consistency is solved, high-quality and high-efficiency generation of the electronic note collage layout is achieved, and the information processing efficiency and cognition experience of a user are remarkably improved.

This invention belongs to the field of visual training and discloses a dynamic visual training method, system, electronic device, and storage medium based on naked-eye 3D imaging. The method includes: acquiring real-time eye movement information of a user and analyzing the real-time movement information to obtain the user's viewing angle change parameters; based on the viewing angle change parameters, using a disparity estimation algorithm to extract and match features from input left and right view images to obtain pixel-level virtual views; dynamically rendering based on the pixel-level disparity map to obtain a 3D view; generating a training task, selecting the 3D view based on the training task, and training the user's dynamic vision. This invention adopts a modular architecture design for artificial intelligence algorithms, supports online learning and updates, and realizes intelligent and personalized visual training, which can effectively improve the user's visual cognitive ability.

Provided are a chimeric protein of two types of rhodopsins, an ion-transporting rhodopsin and a G protein-coupled receptor rhodopsin, and a nucleic acid encoding the same, for the prevention and suppression of progression of retinal diseases, the improvement in visual cognitive behavioral functions (e.g., improvement in light-dark determination functions, improvement in bright spot evading functions, and / or crisis avoidance functions), and the enhancement of visual functions (e.g., improvement in visual acuity). The present invention also provides a method for preventing or suppressing the progression of a disease, disorder or symptom of the retina, for improving a visual cognitive behavioral function (e.g., improvement in light-dark determination functions, improvement in bright spot evading functions, and / or crisis avoidance functions), or for enhancing visual functions (e.g., improvement in visual acuity), in a subject, where the method comprises the step of administering an effective amount of a nucleic acid encoding a chimeric protein of an ion-transporting receptor rhodopsin and a G protein-coupled receptor rhodopsin to the subject.

The application relates to a visual cognitive evaluation method for automatically generating images based on a multimodal language model. The method comprises the following steps: first, an initial image group is used to form an image recognition database, a visual language model is used to expand the image recognition database, then, corresponding images are selected from the image recognition database to form test questions based on a current evaluation task, then, a voice prompt is generated by using a voice generation module based on the test questions, a subject is subjected to a visual cognitive test, and finally, a visual cognitive test result is input into a visual language model to generate a visual cognitive evaluation report. Through the VLM visual language model, accurate image descriptions are automatically generated, and high-precision similar semantic images with great visual differences but small semantic differences are generated by using a semantic fine-tuning technology, dynamic expansion of an image knowledge base and accurate evaluation of subtle differences in visual semantics are realized, and the quality and richness of the content of the knowledge base are improved.

A retinomorphic array is used to convert visual information into electrical signals, and the neural network performs information processing on the input electrical signals to obtain the result of visual cognition; the perception and synchronous preprocessing of visual information is achieved through the retinomorphic array, avoiding the transmission of a large number of redundant visual information from the photoreceptor end to the image information processor, saving bandwidth resources, and improving the efficiency of visual information processing; the use of the crossbar array allows the configuration of a neural network with a more complex structure and more diverse functions, and the higher-level processing of visual information by the neural network realizes a novel neuromorphic vision system integrated therein with image recognition, dynamic tracking, and trajectory prediction.