265 results about "Noise (video)" patented technology

The invention relates to the technical field of biomedical engineering and computer vision, in particular to a non-contact physiological signal extraction method and system based on frequency self-adaption and illumination noise perception.The method comprises the following steps of multi-mode video stream collection and spatio-temporal data preprocessing, illumination-noise perception mask generation and feature filtering, multi-mode video stream collection and spatio-temporal data preprocessing, illumination-noise perception mask generation and feature filtering, and non-contact physiological signal extraction. Frequency adaptive gating and frequency domain feature enhancement, depth time attention feature re-calibration, physiological signal regression and closed loop optimization; the method has the beneficial effects that a lightweight end-to-end deep learning network architecture is constructed by systematically fusing three core modules of illumination-noise perception mask, frequency adaptive gating and depth time attention, and the defects that a traditional physical model depends on artificial prior and is poor in anti-interference performance and high in reliability are overcome. And the one-sidedness caused by high calculation complexity and difficulty in distinguishing the signal and noise of the existing deep learning model is avoided, and the weak physiological signal can be recovered from the face video more accurately and robustly.

The invention discloses a deeply-forged face video frame-level positioning method and system based on weak supervised learning, and the method comprises the steps: firstly constructing a training set with a video as a unit, carrying out the data enhancement of a frame-level sample, and generating an enhanced view pair; secondly, splicing the enhanced view pair and inputting the spliced enhanced view pair into a depth forgery detection model to obtain and generate fusion enhanced frame-level features; then, intra-class contrast learning loss, time sequence consistency constraint loss and frame weight loss are constructed, and a deep forgery detection model is trained based on fusion-enhanced frame-level feature joint optimization. And finally, inputting a video to be detected into the deep counterfeiting detection model to output the frame-level confidence, judging whether the video is a forged video or not, and realizing frame-level counterfeiting positioning. According to the method, video-level detection and frame-level positioning are effectively realized, meanwhile, the influence of label noise in part of forged videos is relieved, and the generalization and robustness of the system are improved.

The invention relates to the technical field of industrial vision and intelligent monitoring, and discloses a loom control system operation state monitoring and fault diagnosis method based on machine vision, which comprises the following steps: acquiring a video stream in a loom shed area and constructing a two-dimensional space-time slice tensor; performing global motion compensation processing on the space-time slice tensor by using a homography transformation matrix, mapping a compensated dynamic texture feature sequence to a three-dimensional phase space by using a time delay embedding algorithm, and reconstructing a closed phase space trajectory representing periodic operation logic of the loom; the discrete Frechet distance between the phase space trajectory of the current operation cycle and the preset reference trajectory is calculated, and a control instruction is generated. The health degree of the sequential logic of the system is directly quantified on the premise that specific components are not recognized by using the invariant characteristic of the phase space manifold topology; the technical problems that small phase lag is difficult to perceive and nonlinear faults cannot be early warned in a strong noise environment are solved.

Methods for motion-controllable video diffusion include extracting optical flow fields from an input video and computing warped noise by iteratively warping noise between consecutive frames using the optical flow fields. The iteratively warping includes (i) re-Gaussianizing expanded pixel regions by sampling fresh Gaussian noise, and (ii) aggregating contracted pixel regions by merging noise particles and renormalizing variance to preserve spatial Gaussianity. An output video is generated by initializing a diffusion process with the warped noise and iteratively denoising to produce temporally coherent output frames. Various other methods, systems, and computer-readable media are also disclosed.

The invention provides a video content enhancement method for a low-illumination environment, and the method comprises the steps: achieving the data preprocessing based on an original low-illumination video frame sequence through frame synchronization, color space conversion and local brightness analysis, generating a noise sensitivity thermodynamic diagram through multi-feature unsupervised learning, and constructing a noise perception gating mechanism through the combination of affine transformation. Dynamic modulation of the characteristic channel is realized; in the multi-scale network structure, a channel attention module is used for carrying out layer-by-layer self-adaptive adjustment on a noise sensitive area; a basic illumination image and an edge enhancement image are generated through double-branch decoding, and then weighted fusion is carried out in combination with a noise thermodynamic diagram, so that brightness balance and detail enhancement are realized; a noise smoothing regular term is introduced during end-to-end training, so that the network achieves dynamic balance between an enhancement effect and noise control.

The invention relates to the technical field of industrial visual intelligence, in particular to an intelligent image selection and switching system fusing visual features and quality scores, which comprises the following steps: receiving multiple paths of video coding streams, inter-frame motion vectors and camera parameters; generating a macro block activeness distribution map based on the video coding stream and the inter-frame motion vector, and performing local window positioning and feature reconstruction on the video coding stream to generate an enhanced video vector; calculating a confidence coefficient mean value and a consistency score of the video coding stream according to the enhanced video vector, and fusing the confidence coefficient mean value and the consistency score with a channel transmission signal-to-noise ratio and a quantization noise increment to generate a video quality score; constructing a multi-criterion optimization model, and setting a feature representation vector for the multi-criterion optimization model; mapping the viewpoint weight based on the inner product of the feature representation vector, and calculating with the video quality score to generate a video switching score; and performing priority ranking based on the video switching score, and triggering a mapping switching instruction. And realizing video image scheduling by fusing the visual features and the quality score.

The invention relates to the technical field of image communication, in particular to a video slow-action frame insertion playback system based on a generative adversarial network, which comprises a video signal acquisition module, a signal processing module, a frame rate conversion module and a playback code output module. The acquisition module outputs an original video frame sequence through an annular buffer; the processing module separates the denoised detail components and the edge gradient features in parallel; the frame rate conversion module uses a geometric structure as a boundary locking condition to guide detail components to execute nonlinear motion compensation and intermediate frame reconstruction; and the playback module executes time base remapping and distributed coding transmission. According to the invention, through a depth generation architecture of structural constraint textures, in combination with space-time consistency verification and a nearest neighbor pixel backfilling mechanism and a virtual time base rate decoupling technology, real-time slow-action video redisk with high signal-to-noise ratio and no artifacts for a high-speed moving target is realized.

The invention discloses a screen content video quality evaluation method and device based on frequency-space complementation and semanteme, and relates to the field of computer vision, and the method comprises the steps: S1, extracting a video block and a key frame of a screen content video, and inputting the key frame into a high-frequency structure texture information extraction branch to obtain high-frequency structure texture information; s2, inputting the key frame into a noise sensing module to obtain a noise sensing feature; s3, inputting the noise perception features into a self-adaptive time sequence embedding module to obtain noise and semantic information; s4, splicing the high-frequency structure texture information and the noise and semantic information, performing quality regression to obtain a quality score of a single-frame key frame, and summing and averaging to obtain a spatial domain video quality score; s5, inputting the video blocks into a Fast-VQA-based quality evaluation branch to obtain a time sequence distortion perception degradation score; and S6, dynamically fusing the spatial domain video quality score and the time sequence distortion perception degradation score to obtain a final video quality score. According to the method provided by the invention, the screen content video quality is effectively evaluated.

The invention discloses a space-time consistent video depth completion method under a zero sample unified diffusion framework. The method comprises the following steps: constructing a depth completion model comprising a variational auto-encoder, a semantic coding network and a space-time diffusion generation network; preparing training data, and generating a frame-level semantic feature vector and a conditional latent variable fusing an original depth and a relative depth for a video frame; training the space-time diffusion generation network in stages by taking the conditional latent variable sequence as input and the semantic features as conditions; in the inference stage, a video sequence to be complemented is processed through a sliding window fusion mechanism, a de-noising depth latent variable is obtained through a trained network, and finally a complemented depth sequence is output through decoding and scale recovery of a variational auto-encoder. The method has the advantages that a depth sequence with measurement consistency, structural integrity and time stability can be generated when depth completion is performed on a sparse, noisy or structurally damaged long sequence video.

Apparatuses, systems, and techniques to encode sequence data in one or more neural networks. In at least one embodiment, a video frame sequence is generated using a neural network to map noise frames to video frames.

The embodiment of the invention discloses a three-dimensional model sequence generation method and related equipment. The related equipment can comprise a three-dimensional model sequence generation device, electronic equipment, a computer program product and a computer readable storage medium. According to the embodiment of the invention, feature extraction is carried out on video frames in a monocular video to obtain image features, an initial noise sequence corresponding to a three-dimensional model of a target object is generated, denoising is carried out on the initial noise sequence according to the image features to obtain a feature sequence set, and based on the frame positions of the video frames and the time distance between the video frames, the target object is obtained. Screening at least one reference feature block associated with the feature block from the feature sequence set, denoising the feature block according to the reference feature block to obtain a target feature sequence of the video frame, and generating a three-dimensional model sequence of the target object based on the target feature sequence; according to the scheme, the reference feature blocks can be screened to perform block-level cross-frame information interaction, so that the generation quality of the three-dimensional model sequence can be improved.

The invention discloses a target image recognition and target detection method based on a video enhancement algorithm, and relates to the technical field of image processing. The method comprises the following steps: firstly, receiving a rain, snow and fog scene video stream through a visual sensor, extracting a video frame target image, performing video enhancement processing, eliminating rain and snow shielding, fog blurring and noise, and generating an effectively enhanced image which is complete in target contour, clear in details and adaptive to subsequent detection; inputting the image into an improved YOLO model of the rain, snow and fog scene, completing feature extraction and category recognition through an optimized feature extraction network, outputting a preliminary target bounding box and a category label, and judging whether a target to be detected and a specific category exist or not; and finally, if the target exists, counting the detection data and carrying out validity verification, thereby realizing high precision, low misjudgment and strong real-time performance of target detection in severe weather of rain, snow and fog, and further effectively solving the problem of high detection result misjudgment rate caused by parameter adjustment lag of adaptive filtering in the prior art.

The present disclosure relates to systems, methods, and non-transitory computer-readable media that generates spatial-temporal positional encodings. For example, the disclosed systems generate a noised token from adding noise to an embedding of a frame of a video. Moreover, the disclosed systems generate a spatial embedding for a token using a centered two-dimensional coordinate map. Further, the disclosed systems generate temporal embeddings for the token from a timestamp of the token in the video. Further, the disclosed systems generate a denoised token by removing noise from the noised token according to spatial-temporal positional encodings that include the spatial embedding and the temporal embedding via a diffusion model. Additionally, the disclosed systems modify parameters of the diffusion model based on a comparison of the denoised token and the token.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating an output video conditioned on an input. In one aspect, a method comprises receiving the input; initializing a current intermediate representation; generating an output video by updating the current intermediate representation at each of a plurality of iterations, wherein the updating comprises, at each iteration: processing an intermediate input for the iteration comprising the current intermediate representation using a diffusion model that is configured to process the intermediate input to generate a noise output; and updating the current intermediate representation using the noise output for the iteration.

The invention relates to a recorded video authenticity detection system and method, and belongs to the technical field of video detection, and the system comprises a client or a front end which is used for obtaining key frame feature setting sent by a back-end server, determining a key frame based on the key frame feature setting and carrying out feature noise injection in a video recording process, after the video recording is finished, the recorded video is packaged, and the packaged video is sent to a back-end server; wherein the key frame feature setting comprises a feature vector setting mode, and the feature vector setting mode comprises a static setting mode and a dynamic setting mode; and the back-end server is used for receiving the packaged video sent by the client or the front end and verifying the packaged video to obtain a verification result, and the verification result comprises a real recorded video or an unreal recorded video. According to the system, the video recorded by the client can be efficiently verified, and the verification accuracy is improved.

The invention provides a production line intelligent conveyor belt autonomous deviation correction method and system, and the method comprises the steps: inputting a video stream into a frame number interpolation module, and carrying out the frame number interpolation of the input video stream; the video stream after frame number interpolation, the historical operation log text and the time series data are input into a lexical element compiling module, text lexical elements are generated from the historical operation log text, sensing lexical element groups are generated from the time series data, and video feature query lexical elements are generated from the video stream; all lexical element information is input into a frozen large language model, and multi-modal semantic high-dimensional features are obtained through analysis according to the advantages of the large language model in the aspects of multi-source data fusion and semantic understanding; and inputting the high-dimensional features into a stream matching action generation module, constructing a vector field consistent in the whole space, and gradually generating a control instruction of a driving motor of a deviation rectifying device matched with the current environment from an initial noise vector under the guidance of the multi-modal semantic high-dimensional features obtained by analysis of a large language model. The conveying belt can be automatically rectified.

The application relates to a video processing method and device, computer equipment and a storage medium. The application relates to the field of artificial intelligence technology, and the method comprises the following steps: acquiring target video information; extracting feature images of each single-frame picture information, and identifying target image edge information in each feature image through a picture edge identification strategy; adding image disturbance noise to image information corresponding to each target image edge information to obtain a plurality of processed single-frame picture information, and combining the processed single-frame picture information into processed picture information; adding audio interference information to audio information, and performing sound elimination processing on target audio data in the audio information that meets a sound elimination processing condition to obtain processed audio information; and determining processed video information according to the processed audio information and the processed picture information. The method can improve the retention degree of effective information of the processed video.

In a case that a scaling value of neural network filter strength is set equal to 0, a decoded image before a deblocking filter is output, and thus block noise occurs. In a case that luma and chroma parameters are not taken into consideration, processing for an image of multiple transfer functions and chroma parameters is not appropriately performed. In a unit of a prescribed block, by using a parameter indicating a degree of application of the NN filter, an image after the deblocking filter and an image after an NN filter are combined using a different ratio of the images. Filter processing of a luma image is switched based on a luma parameter, and filter processing of a chroma image is switched based on a chroma parameter.

The application discloses a kind of based on edge identification no target cable force identification method, the method includes: cable video acquisition and pre-processing: acquisition cable video, video frame is carried out image gray conversion and is adapted to the multi-scale Gaussian filtering operation of filter scale according to noise characteristics;Canny algorithm cable edge identification: using Canny edge detection algorithm to the video frame after pre-processing is carried out cable edge identification;Cable feature point screening and KLT optical flow method identification: from edge identification result screening cable feature point, and using KLT optical flow method in continuous video frame between tracking feature point dynamic displacement;Fundamental frequency identification and cable force calculation: displacement data is applied Fourier transform to identify cable fundamental frequency, and adopts XGBOOST regression model based on fundamental frequency and relevant parameter calculation cable tension.The application realizes the precise, efficient and non-contact monitoring of bridge cable stress state.

This application relates to the field of monitoring data processing technology, and particularly to a method, device, and storage medium for preventing monitoring data from being tampered with. The method includes: during the recording of monitoring data, simultaneously acquiring a thermal noise segment from an imaging sensor for each video frame, reducing equipment costs and deployment difficulty; calculating the image activity metric of the video frame and the noise energy metric of the thermal noise segment, requiring minimal computation and meeting the simple real-time needs of ordinary monitoring equipment; performing an arithmetic combination of the image activity metric and the noise energy metric to obtain a frame verification feature value, which is then associated with and stored with the video frame, making it difficult to forge and having low storage overhead; acquiring the video frame to be verified and generating a verification feature value based on the video frame; determining that the monitoring data has been tampered with when the deviation between the verification feature value and the stored verification feature value exceeds a preset range, thus achieving low-cost, low-complexity, and high-real-time monitoring data integrity protection.

The invention relates to the technical field of computer vision and artificial intelligence security, in particular to a multi-scale 3D field speaking face generation defense method, system, equipment and medium, which actively resist the risk of privacy abuse caused by a speaking face generation technology and solve the problems of visual quality loss, insufficient robustness and the like in the existing defense technology. Based on a space-frequency domain mixed attention mechanism, the method specifically comprises the following steps: firstly, decomposing an input video into a frame sequence, performing significance weighting on noise in a frequency domain, and inversely transforming the noise back to a spatial domain for screening to obtain basic defense noise; performing multi-round iterative optimization on the basic noise through a projection gradient descent loop; and finally, the consistency of noise on a time axis is ensured through a bidirectional time sequence optimization strategy, and the robustness of the noise is enhanced by adopting anti-purification optimization, so that lip motion rendering of the generation model can be effectively interfered, dynamic abnormality of a forged video is generated, and the identifiability of deep forged contents is improved on the premise of keeping the quality of the original video.

The invention discloses a monitoring video delay sweep generation method based on scene change self-adaptive frame extraction, and belongs to the technical field of video monitoring and streaming media processing. The method comprises the following steps: acquiring a sweep strategy containing a time period, a reference interval, a difference threshold value, a forced retention period, a window length and a quality threshold value; receiving a real-time code stream or reading a storage side key frame intermediate file; calculating an inter-frame difference degree by using a brightness channel of a multi-scale structure similarity algorithm, determining a frame extraction interval in combination with a static duration, and marking candidate frames; performing quality detection according to definition, brightness, contrast, stripes and noise, and rejecting unqualified products; when a forced period is reached, a bottom frame is reserved; and performing segmented coding according to a time window, and performing assembly output at the end of a time period. According to the method, online and offline processing links are unified, key information is reserved, and redundancy and time delay are reduced.

In various embodiments, when reference picture resampling is applied to achieve better coding efficiency, methods and devices adapt temporal filtering of a noise-reduced input video prior to encoding according to a resolution ratio. According to one embodiment, temporal filtering strength values are fine-tuned at a pre-processing module according to the resolution ratio of the RPR mode. According to another embodiment, temporal filtering is deferred after rescaling when the RPR mode is enabled, and filtering strength values are fine-tuned according to the resolution ratio of the RPR mode. A third embodiment proposes to refine motion estimation in the encoder by setting the block size of motion estimation in temporal filtering according to the resolution ratio of the RPR mode.

A video coding method and device using an in-loop filter adaptive to various types of noise and characteristics. The video decoding device generates an output block by inputting a reconstruction block for a current block to a deep learning-based in-loop filter. The video decoding device selects a block for retaining the in-loop filter from the reconstruction block and retrains the in-loop filter using the selected reconstruction block and a target block.

The invention relates to a video processing method and device, computer equipment and a readable storage medium. The method comprises the following steps: acquiring continuous original image data frames in a low-illumination environment, and acquiring original noise distribution of the original image data frames; performing noise reduction processing on the original noise distribution to determine noise reduction features in the original image data frame; performing color enhancement processing on the original image data frame to determine a color enhancement feature in the original image data frame; carrying out adaptive feature fusion on the noise reduction features and the color enhancement features to obtain a fusion feature map; performing time domain alignment on adjacent frames before and after the target frame to the target frame to obtain an alignment feature map; and performing pixel enhancement processing on each frame of image in the alignment feature map to obtain a pixel enhancement frame corresponding to each frame of image so as to determine a target video. Therefore, the brightness of the low-illumination video is improved, the noise is remarkably suppressed, the color authenticity is enhanced, the time sequence stability is kept, and the quality and the real-time performance of the video are improved.

The application relates to the technical field of video understanding, and particularly discloses a weakly supervised video anomaly detection method and system based on denoising and deviation removal, which comprises the following steps: calculating anomaly scores of positive and negative package instances, screening top-k instances and constructing instance pairs; calculating loss values of the instance pairs, and dividing the instance pairs into a low-loss reserved set and a high-loss candidate noise set; performing secondary evaluation of the instance pairs in the high-loss candidate noise set in the semantic level by using a pre-trained visual language model, identifying and recalling abnormal samples that are wrongly filtered, and obtaining a recall set; and jointly training the low-loss reserved set and the recall set to obtain an anomaly detection model used for video frame-level anomaly event positioning. Through the dynamic denoising stage and the deviation removal stage based on the visual language model, the application effectively reduces the influence of noise segments on the training process, recovers difficult abnormal samples with discriminative value, and improves the stability and robustness of anomaly positioning.

The invention discloses a video playing processing method and device and a storage medium, and belongs to the technical field of computers. The scheme is applied to an operating system of target electronic equipment and is used for solving the problem of inaccurate video playing state detection in the prior art. The core of the scheme is to call an image display synthesis service of a system bottom layer, obtain frame number feature data in an equipment image display process, and further calculate an average frame rate in unit time. And judging whether the average frame rate is continuously higher than a preset threshold value and reaches a preset duration so as to judge whether the equipment is in a video playing state. According to the scheme, the underlying service of the system is directly accessed, so that non-intrusive acquisition of the frame rate characteristic data from a global perspective is realized, and a detection blind area caused by shielding of an application layer is effectively avoided. Meanwhile, based on the characteristic that the frame rate is continuously high and stable during video playing, transient high-frame-rate noise scenes such as interface sliding and the like can be effectively distinguished, so that the accuracy and robustness of video playing state recognition are improved.

An adaptive object recognition system utilizes correlation-based weighting and scoring that can assist in accuracy and stability. The system receives a plurality of recognition results of an object across a video frame. Weights are assigned to the results based on the stability of the results over the frame and the correlation scores of the neighboring regions. This allows more confidence to be placed in consistent detection and clearer image regions. The system decouples the high accuracy region and the low accuracy region to apply appropriate thresholding. Time analysis is used to support frequently detected objects. The weighted results are combined to generate a final recognition that is more robust to non-uniform image quality and noise.

Aspects of the present disclosure relate to providing a shoppable video corpus by generating a shoppable URL. A URL is extracted from a video corpus, in which the extracted URL may be a long URL or a short URL. Extracted URLs are then combined and normalized. From the normalized URL, noise is removed and quality control is performed. As a result, shoppable URL may be presented at the user’s computing device as personal recommendation. The video and metadata of the cleaned URL is also ingested and stored in a database for future reference.

A video recommendation method fusing social information comprises the steps that firstly, a user-video interaction graph and a social graph are constructed through video media platform data, different combination strategies of a diffusion model and graph convolution operation are applied in a denoising priority path and a structure priority path respectively, and collaborative perception and fusion of user embedding and structure information are achieved; secondly, aligning user embedding output by the double-track denoising path by adopting comparative learning, and enhancing robustness; secondly, performing deep coding on the user-video interaction graph and the social graph by applying a double-graph neural network, comprehensively capturing user preferences and embedding videos; and finally, for noise possibly introduced by multi-module fusion, deep denoising is executed to purify user embedding again. According to the method, guidance of structural information is introduced in the diffusion denoising process, and more robust user embedding is obtained by using a comparative learning strategy, so that a structural perception high-fidelity social denoising task is completed, and personalized video recommendation is realized on a video media platform.