67 results about "Video detection" patented technology

This invention discloses a method and apparatus for detecting deepfake videos based on the consistency of emotion change rates. Applied to the field of data processing technology, this invention is a deepfake video detection scheme that achieves authenticity determination by extracting and comparing the emotion change rates of faces and audio through multiple steps. First, OpenCV is used to read video metadata, and librosa is used to decode audio, unifying the frame rate and sampling rate. Then, OpenCV is used to detect faces and generate continuous time segments. ResNetFaceEmotion is called to extract facial emotion features, and the facial emotion change rate is obtained through sliding windowing, noise reduction, and first-order difference. Simultaneously, EnergyVAD is used to filter and segment the audio, and features are extracted using a preset model to obtain the audio emotion change rate. The intersection of face and audio time segments is taken, and time alignment is achieved using the DTW algorithm and Euclidean distance. Finally, an evaluation model containing bidirectional GRU, self-attention pooling, and a fully connected network is constructed, with the aligned features as input and the detection results as output.

This invention discloses a robust detection method for AI-generated videos based on gradient-guided contrastive learning, comprising the steps of: constructing and training a spatial feature extractor; using the trained spatial feature extractor to train a temporal feature extractor and a classification head; inputting the video to be detected into the trained spatial feature extractor, temporal feature extractor, and classification head to obtain the probability that the video to be detected is an AI-generated video; and obtaining the detection result according to a set detection threshold. The beneficial effects of this invention are that by employing a gradient-guided curriculum learning strategy to construct color-enhanced video frame samples, and by using a supervised contrastive learning method to train the spatial feature extractor and the frame sequence samples to train the temporal feature extractor, the robustness of the AI-generated video detection method to color-related post-processing can be improved; and by using a curriculum learning strategy to dynamically adjust the intensity of color enhancement during training, higher training stability and robustness are achieved.

The application provides a bullet cabinet with alcohol detection function, and the alcohol detection system comprises: a data acquisition module which acquires detection video of a cabinet opening personnel through a camera, acquires detection voice of the cabinet opening personnel through a microphone, acquires historical cabinet opening data, and acquires current time data; a personnel state determination module which processes the detection video, the detection voice, the historical cabinet opening data, and the current time data to obtain a face alcohol state score, an eye state abnormality score, a body action amplitude score, a voice instability score, and a time risk score; a risk index determination module which determines a final risk index according to the face alcohol state score, the eye state abnormality score, the body action amplitude score, the voice instability score, and the time risk score; and a cabinet lock control module which controls the opening and closing state of an intelligent cabinet lock according to the final risk index. The alcohol detection precision of the cabinet opening personnel is improved, and the risk of using a gun by the cabinet opening personnel is improved.

Embodiments of the present application provide a video detection model training method and device, a video detection method and equipment. The video detection model training method comprises: obtaining an original data set; based on a preset hierarchical artifact classification system, each training video in the original data set is labeled to obtain the label of each training video, wherein the label comprises artifact type, time period, coordinate information and thought chain information; according to each training video and the label corresponding to each training video, a training data set is constructed; based on the training data set and a first preset training function, a pre-trained video detection model is trained for the first time to obtain a first video detection model; based on the training data set and a second preset training function, the first video detection model is trained for the second time to obtain a target video detection model. Thus, the accuracy, robustness and interpretability are improved, laying a high-security scene reviewable, acceptable AI video authentication baseline.

A video system comprises a video overlay processor. The video overlay processor is configured to receive a video stream from a camera system in real time during operation of a platform. The video overlay processor is configured to overlay a first pattern and a second pattern on an alternating and repeating basis onto frames sequentially in the video stream in real time to form a modified video stream. The video overlay processor is configured to display the modified video stream in real time on a display system, wherein the first pattern and the second pattern in the frames in the modified video stream are perceptually hidden on the display system in response to the frames in the modified video stream being displayed without interruption.

This invention discloses an intelligent diagnostic system and method for intersection operation status based on feature matching. Specifically, it involves: collecting peak and off-peak video data from an unmanned aerial vehicle (UAV) above the intersection; detecting dynamic parameters such as traffic flow and speed of motor vehicles and non-motor vehicles at the intersection, as well as static parameters such as the coordinates of lane markings and guide arrows inside and at the intersection's entrances; constructing a multi-dimensional fusion indicator system; establishing a database of intersection influencing factors and a database of improvement measures; simulating and extrapolating data obtained from static modeling and dynamic perception using digital twin technology to acquire data features of different traffic scenarios from the intersection influencing factor database; and achieving a one-to-one strong rule association between the influencing factor database and the improvement measure database through multi-dimensional Apriori association rules. This invention effectively improves the low accuracy and efficiency of existing non-motor vehicle detection methods and provides adaptive diagnostic evaluation for intersections of different specifications.

Provide a system. 【Solution means】 Means for receiving video data from a video acquisition device, Means for analyzing the video data to detect an individual, Means for generating and superimposing a virtual display corresponding to the individual, Means for streaming the virtual-displayed video to an information device, Means for detecting abnormal behavior from the video and generating a notification, Means for canceling the virtual display when a specific condition is satisfied, Means for transmitting the notification of the abnormal behavior to a mobile information device and displaying it to a user, A system including the above.

The application relates to a rule violation video detection method and system based on skeleton behavior recognition, which comprises a video acquisition module, a key frame extraction module, an image data set establishment module and a real-time video behavior detection module.The posture estimation algorithm is used in the video key frame extraction module to estimate the joint position, the rule violation behavior detection module adopts a multi-filter dynamic graph convolutional neural network model, the real-time video behavior detection module adopts a multi-filter dynamic graph convolutional neural network, and the connection relationship between different behavior nodes is adapted through a dynamic skeleton graph.The Inception structure and the dynamic skeleton graph method based on the graph convolutional neural network can realize real-time and accurate detection and identification of rule violation sensitive action behaviors in internet live broadcast, thereby providing an important guarantee for standardizing the anchor live broadcast behavior and providing a civilized internet live broadcast environment.

A video inspection method and apparatus. The method comprises: on the basis of a video frame sequence corresponding to a target video, obtaining an environment image sequence; acquiring target sensor data corresponding to the target video, the target sensor data being provided by a target sensor configured in a target device that acquires the target video, and the target sensor being used for sensing motion of the target device; and, on the basis of the environment image sequence and the target sensor data, determining whether the target video is a fake video.

This disclosure provides a video processing method, apparatus, device, and storage medium, relating to the field of artificial intelligence technology, and applicable to various scenarios such as cloud technology, artificial intelligence, and automotive applications. The method includes acquiring a video to be processed, which comprises multiple consecutive image frames; determining a multi-layer image corresponding to each image frame; extracting color features from the multi-layer image corresponding to each image frame to obtain color feature information corresponding to each image frame; and for any two adjacent image frames, performing flicker detection on the two image frames at at least one feature scale based on the color feature information corresponding to the two image frames, thereby reducing reliance on manual review and improving video detection efficiency and accuracy.

The application relates to the technical field of camera video processing, in particular to a camera video processing method, device and medium. The method comprises the following steps: acquiring the state of a monitoring area corresponding to each camera in a target camera set in a target time period; the state of the monitoring area corresponding to any camera in the target time period is obtained based on preset type data of the monitoring area corresponding to the camera in the target time period, and the preset type data does not include the video of the camera; if the state of the monitoring area corresponding to a certain camera in the target time period is abnormal, a preset video processor is used to process the video collected by the camera in the target time period; and if the state of the monitoring area corresponding to a certain camera in the target time period is normal, the preset video processor is not used to process the video collected by the camera in the target time period. The application can process the video in a targeted manner, reduce the waste of computing power and reduce the video detection cost.

The application discloses a dense video detection method based on a large-scale language model, which comprises the following steps: collecting an existing image caption dataset and a video caption dataset, and constructing image-text pairs and frame image sequence-text pairs; building a multi-modal large language model framework, performing distributed training through a DeepSpeed tool, and fine-tuning the multi-modal large language model by using the image-text pairs and the frame image sequence-text pairs; collecting an existing dense video dataset, uniformly sampling frames of the dense video to obtain frame image sequences, and filling and truncating events of the dense video; and constructing a model input paradigm, and performing supervised fine-tuning on the fine-tuned multi-modal large language model to realize dense video detection. The application combines dense video detection with a multi-modal large language model, and fully utilizes text information of image caption and video caption tasks and visual and text understanding capabilities of the multi-modal large language model through transfer learning.

The application discloses a digital person video detection method based on facial semantic features, acquires real person voice visual video to constitute a training data set; constructs a digital person video detection model based on facial reality mode representation, respectively extracts facial reality mode representation features, mouth features and voice features, after multi-modal feature fusion, generates a matching score matrix according to the fused features; trains the digital person video detection model by using the training data set, generates a matching score matrix by using the trained digital person video detection model, and aggregates the matching score matrix to obtain the overall matching score of the to-be-detected video, so that the forgery detection is realized. The application proposes a unified audio and video forgery detection framework based on the facial reality mode, and improves the deep forgery detection performance in different scenes.

The application discloses a video detection method and system based on spatial domain and time domain characteristic cable frequency, and belongs to computer vision technology. The method comprises the following steps: inputting an image sequence; extracting phase information, including extracting phase information of a detection object and a background in a phase space; filtering out background phase, including calculating local spatial feature information after an adaptive direction filter to screen out the phase of the detection object and the background, filtering out the phase information of the background, and reserving the phase information of the detection object; and extracting a vibration signal and a frequency, including extracting a frequency value of vibration of the detection object. Compared with the prior art, the application has the advantages that weak vibration signals and vibration frequencies of a cable structure can be well extracted in a complex background, so that the cable force of the cable can be accurately calculated.

The application discloses a counterfeit video detection method, device, equipment, medium and program product. The method comprises the following steps: acquiring at least two target image frames of a to-be-detected video; wherein the target image frames are obtained by screening based on the differences between different image frames in the to-be-detected video; acquiring image description information of the target image frames; performing a preset each authenticity detection task on target detection data of the to-be-detected video to obtain a detection result corresponding to each authenticity detection task; wherein the target detection data at least comprises the image description information; and determining a counterfeit detection result of the to-be-detected video based on each detection result. The application can identify the authenticity of the content of the to-be-detected video, avoid the training set coverage limitation caused by recognizing the visual attributes based on the face and the picture, has better generalization, and can further improve the accuracy of detecting various counterfeit videos.

The application relates to the technical field of terminals, and discloses a video detection method, an electronic device, a storage medium and a program product. In the video detection method, the electronic device can acquire a call video, acquire a plurality of frames of to-be-detected images from the call video, and perform AI face changing risk detection based on the plurality of frames of to-be-detected images. In the detection process, the electronic device can perform forgery detection based on the Nth frame of to-be-detected images after performing face detection based on the Nth frame of to-be-detected images, and in the process of performing forgery detection based on the Nth frame of to-be-detected images, face detection is performed based on the (N+1)th frame of to-be-detected images, that is, the face detection process and the forgery detection process can be executed in parallel. In this way, the time consumption of the detection process of the plurality of frames of to-be-detected images can be shortened, and the detection efficiency is improved.

The challenge lies in how to distinguish and detect each group of microparticles when multiple groups of microparticles are present. [Solution] The microparticle swarm detection device 10 irradiates a predetermined area including the workpiece 60 with green sheet laser light and red sheet laser light, switching the emission timing of these lights in a time series, and uses the imaging unit 15 to capture a video of the predetermined area including the workpiece 60. Then, frames captured with green sheet laser light and frames captured with red sheet laser light are extracted from the captured video to generate a video of green frames and a video of red frames. Subsequently, the microparticle swarm detection device 10 detects microparticle swarms based on the extracted video of green frames. The microparticle swarm detection device 10 also detects microparticle swarms based on the extracted video of red frames.

This invention relates to the field of computer vision technology and discloses an AI-synthesized video detection method and system based on multi-agent collaboration. The method includes: acquiring the video to be detected and sampling keyframes; understanding the keyframes based on a visual language model and generating a forgery hypothesis for the current round based on the generated global semantic context; generating a spatial-temporal routing query; performing target localization and segmentation on the video to be detected based on the spatial-temporal routing query; extracting structured physical evidence from the target region video; performing evidence-constrained reasoning to obtain a preliminary judgment result and confidence level; performing consistency checks on the execution trajectory to obtain a reliability score and feedback information, and outputting the final authenticity judgment result. This invention improves the reliability, generalization ability, and traceability of AI-synthesized video detection by combining contextual modeling, hypothesis-driven routing, multi-dimensional physical evidence collection, evidence-constrained reasoning, and closed-loop reflective verification.

This specification provides a video detection method, apparatus, device, storage medium, and program product. Specifically, multiple video frames are acquired from the video to be detected, and multimodal features of the multiple video frames are extracted; based on the multimodal features of the multiple video frames, a multimodal feature sequence is generated, and word embedding processing is performed on the speech content text of the multiple video frames to obtain a word embedding representation sequence; the multimodal feature sequence is enhanced using a spatiotemporal fusion network to obtain an enhanced multimodal feature sequence; based on the enhanced multimodal feature sequence, the word embedding representation sequence, and video detection task prompts, a multimodal large language model is triggered to determine the video detection result of the video to be detected. This improves the accuracy of video authenticity detection.

The present application relates to a kind of mine chain feeder chain detection method and detection system, and detection system includes computer analysis system, tension detection system, laser radar detection system and video detection system.The method includes the following steps:1 obtains the tension data, texture data and shape deformation data of chain at multiple continuous time points;2 data is normalized to obtain tension feature a, texture feature b and shape feature c;3 construct data set at t time;4 input data set at t time into trained Transform-LSTM model and output the health index h of chain;5 set the health threshold H of chain, if the health index h of chain is less than the health threshold H of chain, then determine that chain is in healthy state.The present application measures tension, texture and shape of chain in real time, and can obtain the safety range and state of chain in real time through comprehensive comparison and judgment.

The application discloses a model training method and device, a nonvolatile storage medium and an electronic device. The method comprises the following steps: acquiring a preset video set with quality defects, wherein the preset video set comprises labeled videos and unlabeled videos; using a preset generator to add pseudo-defect labels to the quality defects in the unlabeled videos to generate pseudo-labeled videos, and extracting defect features in the labeled videos and the pseudo-labeled videos to generate defect feature videos, wherein the defect feature videos indicate defect features through preset defect labels, and the preset defect labels comprise true defect labels and pseudo-defect labels; using a preset discriminator to perform label discrimination on the preset defect labels in the defect feature videos to generate a discrimination result; and using a preset loss function and the discrimination result to train an initial video detection model to generate a target video detection model. The application solves the technical problem of low model training efficiency.

This invention provides a hate video detection method based on a multimodal pattern library and a hierarchical complementary gating mechanism, belonging to the field of multimodal learning technology. The method includes: acquiring textual, visual, and audio features of a video through a multimodal feature extraction module and mapping them to a unified space; constructing a multimodal pattern library using a large language model, refining massive training samples into a compact and interpretable high-level semantic prototype set; injecting semantic priors through a pattern library retrieval and feature enhancement module; employing a hierarchical complementary gating mechanism to dynamically allocate modal weights with text as the semantic anchor, achieving selective fusion of audio and visual features; and finally training the model using a composite loss function to complete hate video detection. This invention solves the problems of modal competition, visual noise sensitivity, semantic ambiguity in instance-level retrieval, and high computational overhead caused by symmetric fusion in existing methods, significantly improving detection accuracy and efficiency, and enhancing model robustness and generalization ability.

The application relates to the technical field of video detection, and discloses a false news video detection method based on information enhancement and guided denoising, which comprises the following steps: obtaining a news video to be detected, and extracting label features of the news video to be detected by using a large language model; based on the label features, multi-modal features of the news video to be detected are extracted, and the label features are subjected to semantic enhancement to obtain label enhanced features; label guided denoising is carried out based on the label enhanced features and the multi-modal features, multi-modal denoised features are obtained, and whether the news video to be detected is a false news video is determined according to the multi-modal denoised features. The application extracts key information of a multi-modal video by means of the semantic understanding ability of the large language model, extracts multi-modal features, subjects the label features to semantic enhancement, combines a label guided denoising mechanism, constructs a complete and effective false news video detection process, improves the detection precision and efficiency of the false news video, and realizes more reliable news authenticity verification.

The application relates to the field of video detection, and discloses a weakly supervised video anomaly detection method, device and equipment and a storage medium. The method comprises the following steps: extracting video features; constructing an anomaly judgment model comprising a multi-scale time feature fusion module, a multi-scale space feature fusion module and a video anomaly score module; training the anomaly judgment model by using the extracted video features; performing feature extraction on a to-be-detected video and inputting the to-be-detected video into the trained anomaly judgment model to perform anomaly judgment. In this way, the convergence degree of the model and the learning range of the model on the video are increased by fusing the multi-scale time and space features of the video, the understanding ability of the model on the video features is improved, sufficient video information can be learned, the model is more robust, the model can better utilize the existing features without changing the number of training data, the video can be automatically judged for anomaly, and the efficiency of video anomaly detection is greatly improved.

The present disclosure provides an AI-powered traffic control system (102) mounted on a vehicle (301) for autonomous monitoring and management of traffic violations and conditions. The system includes at least one AI-powered camera (201) to capture images or video, detect traffic violations, conditions, and unsafe driving behaviours using image recognition and machine learning algorithms; a speed detection device (202) to measure vehicle speed; and an adaptive light system (203) to deliver real-time visual feedback to surrounding drivers. A processor (501) analyses data from the camera (201) and speed detection device (202), controlling the adaptive light system (203) and communicating with external systems. Memory (502) stores data for reporting and post-incident analysis. A method for implementing the system includes initializing components, capturing data, analysing violations, providing feedback, and ensuring privacy protection. This system (102) enables decentralized traffic management, enhancing road safety and regulatory compliance in varied traffic environments.

The present application relates to the field of artificial intelligence and automatic inspection, and particularly relates to a kind of inspection method based on Nanobot multi-agent framework, comprising: configuring multiple agents, the multiple agents include inspection planning agent, video processing agent and data summary agent, each agent is deployed and independently runs in Nanobot multi-agent framework;For at least part of the agent, configure Nanobot tool module, and configure input and output interface for Nanobot tool module;Through inspection planning agent, issue inspection task to video processing agent;After video processing agent receives inspection task, corresponding Nanobot tool module is called to complete video detection, and then inspection information is transported to data summary agent;Data summary agent uploads the inspection information summarized to vector memory module;Based on early warning mechanism, the inspection information summarized in vector memory module is detected, and selective early warning is issued.

This invention discloses a video detection method and apparatus based on multimodal thinking chain and direct preference optimization. The method includes: acquiring a set of video samples with known labels as a training set, constructing enhanced input samples containing multi-source information, and further optimizing an open-source visual language model to generate predicted labels and inferences for unknown videos. By constructing a two-stage framework including information enhancement and inference enhancement, combined with multimodal thinking chain and direct preference optimization algorithms, the method mathematically ensures the model maximizes mutual information utilization of multimodal contextual information and effectively suppresses the interference of spurious relevance through a contrastive learning mechanism. Experimental results show that this invention achieves good detection performance and exhibits significant advantages in interpreting the amount of generated information, logicality, and persuasiveness.

The application provides a harmful video detection method and device based on cross-frame timing semantic understanding. Based on the feature that harmful semantics in a video is often jointly constituted by entity attributes, entity relationship and timing evolution process, a hierarchical analysis and joint reasoning strategy for complex semantic structure is proposed. A hierarchical modeling and joint decision technology is adopted to perform harmful semantic analysis from three aspects of entity attribute and emotion modeling, entity timing relationship modeling and global semantic graph reasoning, and finally determine through semantic fusion, so as to improve the recognition ability of implicit harmful content and rule violation semantics in complex context.