138 results about "Online video" patented technology

The invention discloses an online video content intelligent pushing method combined with a learning interest model. The method comprises the following steps: constructing a dynamic interest vector based on multi-source user behavior data, generating a user interest portrait vector set, and performing interest dimension clustering and weight distribution; generating a video content feature vector set according to a clustering result of the user interest portrait vector set; establishing a multi-dimensional association relationship between the user interest portrait vector set and the video content feature vector set, and outputting a user-video matching confidence matrix; converting the user-video matching confidence coefficient matrix into a push sequence based on a multi-objective optimization strategy and issuing the push sequence; and feedback behaviors of the user on the pushed video are collected in real time to realize closed-loop optimization. The method has the following advantages and effects: accurate perception and deep semantic matching of the dynamic learning interest of the user can be realized, and the accuracy, timeliness and user satisfaction of content distribution are remarkably improved, so that the learning efficiency and experience are optimized.

The invention relates to a weak supervision online video moment positioning method and system based on memory perception, and belongs to the technical field of artificial intelligence, and the method comprises the steps: carrying out the multi-modal feature fusion of a given video and a text query thereof, and obtaining the unified frame level representation at each stage; inputting the fused features into an offline module and an online module in an integral and frame-by-frame manner by using an offline guide online model architecture; in the off-line module, generating a Gaussian mask to reconstruct query of a covered part of words, and obtaining a proposal of an action starting moment; in the on-line module, the long-term historical memory in the window is used for enhancing the score, the attention weight of the score in the window is dynamically generated, and the score of the current frame is calculated in a weighted mode; taking the proposal obtained by the offline module as a pseudo tag, and providing supervision information for the score sequence of the online module; and high-performance weak supervision on-line moment positioning can be completed only by independently deducing the on-line module. The expansion capability and the application value of the model are remarkably improved.

The invention discloses a video conference multi-modal data alignment method and device based on a causal mask, equipment and a medium, and relates to the technical field of computers, and the method comprises the steps: carrying out the feature extraction and fusion of an original audio, an original video stream and an original document in an online video conference, time sequence division is carried out based on the obtained multi-modal fusion features to obtain a triple time sequence window; determining an initial weight value corresponding to the triple time sequence window, and performing normalization adjustment on the initial weight value by using a preset constraint condition to obtain an adjusted weight; indexing a preset time sequence offset matrix by using a speaking identifier of a speaking party, correcting an original time sequence of the triple time sequence window based on an indexing result, and determining a target attention result corresponding to the triple time sequence window by using a preset causal mask mechanism, and performing multi-level alignment fusion on the multi-modal fusion features based on the target attention result to obtain a multi-modal alignment result. The precision of the multi-mode alignment technology is improved, and future information leakage is avoided.

The invention discloses an optimization method for realizing zero delay of video switching, which is particularly used for online video playing and streaming media service. Comprising the following steps: monitoring the residual playing time of a current video through a background thread, starting frame extraction and decoding operations of a next video in advance in the last few seconds of video playing, and caching the frames after parallel decoding to a high-speed cache region. When the video is switched, the timestamps of the current video and the next video are accurately connected through a double-timestamp synchronization mechanism, and the phenomena of black screen and delay are reduced. And dynamically adjusting rendering paths and priorities by adapting to graphic APIs of different hardware platforms. The method and the device are suitable for common video playing, advertisement insertion and real-time video streaming media playing in a complex network environment, and the video switching fluency and the user experience can be greatly improved. Through multi-thread parallel decoding and cache management, utilization of equipment resources is effectively optimized, and especially in a low-end equipment or high-load environment, high efficiency and stability of video playing can still be kept.

The invention provides an online video instance segmentation method and system based on mask propagation, and the method comprises the following steps: S1, achieving the cross-frame propagation of a target mask through a mask propagation model, obtaining a prediction mask of a current frame, and achieving the correlation of a target through the calculation of the intersection-union ratio of the prediction mask; and S2, performing back propagation on the target by using the mask propagation model so as to complement the missed target mask and enhance the continuity of the track. Objects between different frames are associated on the basis of a mask propagation mechanism and in combination with the intersection-to-union ratio of masks, and meanwhile, in order to relieve the scene that a segmentation model fails to segment shielded objects, non-significant objects and the like, a reverse mask propagation technology is provided to complement the failed frames. The method aims at performing fine segmentation and continuous tracking on various scenes such as intelligent monitoring, automatic driving, augmented reality and video editing on the target in the video, and has wide research and application prospects.

Systems and methods are provided herein for only including portions of a user's environment that have been approved by a user in a video conference while excluding portions that have not been approved. This may be accomplished by a device receiving a policy identifying one or more approved objects of a scene of a video stream. The device may then generate a filtered video stream by only including portions of the scene that comprise the one or more objects that were approved by the policy in the filtered video stream. The filtered video stream may be combined with other video streams to generate a video conference that is transmitted and / or stored by one or more devices participating in the video conference.

The invention provides a method for facilitating audio communication between a first user device and a second user device connected via a network during an online video gaming session, the method comprising: receiving first audio data from an audio input device associated with the first user device, the first audio data representing one or more speech samples; generating text data representative of the first audio data; transmitting, by a network, the text data; generating second audio data based on the text data; and outputting audio based on the second audio data at an audio output device associated with the second user device.

Systems and methods are provided herein for only including portions of a user's environment that have been approved by a user in a video conference while excluding portions that have not been approved. This may be accomplished by a device receiving a policy identifying one or more approved objects of a scene of a video stream. The device may then generate a filtered video stream by only including portions of the scene that comprise the one or more objects that were approved by the policy in the filtered video stream. The filtered video stream may be combined with other video streams to generate a video conference that is transmitted and / or stored by one or more devices participating in the video conference.

The invention provides an interactive digital human presentation method based on a backplane video and a related device, and relates to the technical field of artificial intelligence such as human-computer interaction, digital human, end-cloud integration and large models. The method comprises the following steps: extracting a real-time inquiry from a real-time interaction request initiated by a user for a digital human offline video in a playing state, and sending the real-time inquiry and historical interaction content to a cloud server; the cloud server is controlled to generate reply voice according to the real-time inquiry and historical interaction content and select a target digital person bottom plate video matched with the reply voice from a digital person bottom plate video library, and different digital person bottom plate videos correspond to digital persons showing different actions respectively; controlling the cloud server to generate a digital person online video for answering the real-time inquiry according to the target digital person bottom plate video and the reply voice; and linking and playing the digital human online video stream-pushed by the cloud server in the digital human offline video in the playing state in a manner of inserting the waiting transition video.

A method is provided for improving multiplayer user experience in an online video game utilizing a lockstep engine. The method includes receiving user input information from at least one of a plurality of users of a match of the video game that utilizes the lockstep engine. Each of the user input information from the at least one of the plurality of users being associated with a current frame of a plurality of frames of the lockstep engine based on the respective user input information being received within a current time interval of the current frame. The method also includes sending the current frame of the user input information irrespective of whether the user input information is received from each of the plurality of users within the current time interval. Clients no longer need to be in the game at the same time for the game to proceed.

A method including: accessing a first configuration; accessing a primary video stream comprising a first set of video content, from a first online video platform; accessing a secondary video stream comprising a second set of video content; and at an initial time, combining the primary video stream and the secondary video stream according to the default viewing arrangement; at a first time, detecting the first trigger event in the primary video stream; in response to detecting the first trigger event, combining the primary video stream and the secondary video stream according to the first target viewing arrangement, and publishing the first composite video to a second video platform; and at a second time, detecting the second trigger event in the secondary video stream; in response to detecting the second trigger event, combining the primary video stream and the secondary video stream according to the second target viewing arrangement.

The invention discloses a real-time remote guidance system and method based on online video key information fusion. The real-time remote guidance system and method are used for helping other personnel to operate instruments and equipment on line by experts or equipment skilled operators through remote immersive videos. According to the invention, two controllers and camera devices at different geographic positions are utilized, one camera device is used for shooting instruments and equipment needing to be operated by a guided person, the other camera device is used for shooting hand actions of the guided person, and the main controller carries out real-time fusion on contents shot by the two camera devices, so that the real-time fusion of the contents is realized. A fusion result is displayed on a video display device of an auxiliary controller used by a guided person, and clear and visual remote online guidance is provided for an unskilled operator. By adopting the technical scheme of the invention, the learning cost of technicians in operating instruments and equipment is saved to a certain extent, and the communication cost between the instructor and the instructed person is reduced.

The invention provides an interactive digital human presentation method based on air port connection and a related device, and relates to the technical field of computers, in particular to the technical field of man-machine interaction, natural language processing, intelligent robots and other artificial intelligence. The method comprises the following steps: determining a target digital human offline video corresponding to an initial query demand input by a user; in response to the absence of other digital person offline videos matched with the real-time inquiry initiated by the user for the target digital person offline video in the playing state, sending the real-time inquiry, the gas port time data of the target digital person offline video and the historical interaction content to a cloud server; the cloud server is controlled to generate a digital human online video for answering the real-time inquiry and determine a target gas port used for connecting the digital human online video without waiting in the target digital human offline video based on the gas port time data, the real-time inquiry and the historical interaction content; and linking and playing the digital human online video pushed from the cloud server when the target digital human offline video is played to the target gas port. According to the method, high-precision and low-delay man-machine interaction is realized.

The present application relates to a memory-aware based weakly supervised online video moment localization method and system, belonging to the field of artificial intelligence technology, comprising: multi-modal feature fusion on a given video and its text query to obtain unified frame-level representation at each stage; using an offline-guided online model architecture, the fused features are input into the offline and online modules in the form of the whole and frame by frame; in the offline module, a Gaussian mask is generated to reconstruct the hidden part of the query, obtaining the proposal of the action starting moment; in the online module, the long-term historical memory in the window is used for enhancement, and the attention weight in the window is dynamically generated, and the score of the current frame is calculated by weighting; the proposal obtained by the offline module is used as a pseudo label to provide supervision information for the score sequence of the online module; only the online module needs to be inferred separately, and the weakly supervised online moment localization with high performance is completed. The present application significantly improves the expansion capability and application value of the model.

The invention discloses a conference software interaction method and device, equipment and a storage medium, and relates to the technical field of video conferences, and the method comprises the steps: obtaining the flow consumption generated by a current video conference in real time; when the flow consumption exceeds a preset flow threshold value, judging whether the current conference initiating terminal enters an anti-interference mode or not; if yes, generating an interference software list through the current conference initiating terminal; and performing anti-interference processing on the current conference initiating terminal and the target conference participating terminal based on the interference software list, wherein the target conference participating terminal is the conference participating terminal selected to enter the anti-interference mode. By applying the technical scheme, the technical problem that in the prior art, when an online video conference is carried out through conference software, if the network quality is poor, the conference effect becomes poor is solved, and then the user experience of participating in the conference is improved.

An ad player presents ads in association with a video player by evaluating an associated ad script. The ad player transforms data included in the ad script into operational instructions. Hence, the ad player flexibly and dynamically configures itself and presents ads in accordance with the contents of the ad script, enabling a publisher to modify advertising aspects simply by modifying the ad script. The ad script can comprise a script in a tag-based markup language that is readable by the ad player. For example, the ad script can include one or more tags, each tag including one or more attributes that are each set to a value. The ad player determines the values of the attributes and presents ads in accordance with associated ad characteristics or behaviors.

A method for an online video game or application service system includes running a video game or application on an application host server at a data center, an uncompressed video stream being produced therefrom. The uncompressed video stream is encoded into compressed video stream, which is then transmitted over the Internet to an output client device of a user. The output client device decompresses the compressed video stream and displays live video on a screen. User control input transmitted from an input client device is delivered to the application host server. The user control input includes game or application commands. The input client device is associated with the user and is separate from the output client device. Responsive to receiving the game or application commands, the application host server generates a new uncompressed video stream.

The present invention relates to the technical field of fire scene investigation, and specifically to a method for on-site positioning and restoration of a fire point via a surveillance video. The method comprises the following steps: exporting a surveillance video into a video format recognizable by a PC, searching for characteristic frames, recording characteristic frames where flames or smoke first appear, converting the frames into picture A, recording coordinates of the leftmost and rightmost pixels of the flame, connecting the method to an on-site positioning device to collect online video and read the video of a surveillance camera, drawing an on-site position map, determining the closest and farthest possible distances between the flame position at the fire scene and the camera, finding an on-site position corresponding to these coordinate points and marking it in a plane map, and marking the range of the fire point in the video, i.e., the range formed by the projection of all possible positions onto the ground. Compared with the prior art, the present invention calculates and exports the fire coordinate points in the original video, loads the video signal of the video camera at the same position at the scene into the device, locates the fire point at the actual scene, and thus achieves high-precision positioning at the actual scene.

A video conference device includes a camera, a communication circuit and a processor. The camera is configured to capture a real-time video of a first location. The communication circuit is communicatively connected to a remote server and a first electronic device located at the first location. The processor executes an online conference application, and processes the real-time video and real-time visual signals received from the remote server via the communication circuit. The processor executes the online conference application to establish or join an online video conference on the remote server. The processor obtains first authentication information from the first electronic device via the communication circuit, sends the first authentication information for identity authentication, and receives first operation authorizations granted to the first electronic device by the remote server based on the first authentication information. The first operation authorizations enable the first electronic device to control a first representative cursor.

1. Name of the product in this design: Wireless Car Box (CP029-3). 2. Purpose of this design: To enable functions such as mobile phone interconnection and online video through the original car CarPlay channel. 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.

The invention belongs to the field of target tracking, and particularly relates to a DeepSort-based multi-target pedestrian tracking method. According to the invention, the real-time online video is obtained and is preprocessed; according to the invention, a tracking algorithm of DeepSort is optimized on the basis of an improved YOLOV5 detector. Firstly, a YOLOV5 detector is improved, a triple attention mechanism is added to the YOLOV5 detector, the detection precision is improved, follow-up tracking is facilitated, and under the detection condition, detection result information is input into a DeepSort algorithm. Secondly, in DeeepSort, trajectory curve optimization based on context association is provided, and the problem that Kalman filtering cannot be effectively predicted due to the fact that noise of state estimation of Kalman filtering is larger and larger under time accumulation under the shielding condition is effectively solved. And finally, in the tracking matching process of DeepSort, replacing the original IOU matching with CIOU matching to optimize the matching process of target tracking. In a multi-target pedestrian tracking task, accurate tracking can be more effectively carried out.

The present application provides an asynchronous multi-threaded self-service sampling method, server-side equipment, system and electronic terminal, including: obtaining the identity identification information of the person to be tested, thereby obtaining the detection task corresponding to the person to be tested; classifying the detection tasks of the same type into the same detection task queue, and queuing and managing different detection task queues and their corresponding detection tasks through multi-threading; responding to the asynchronous online self-service sampling request information of the person to be tested, supervising and guiding the online self-service sampling process of the person to be tested by conducting online video in the corresponding thread to obtain qualified self-service sampling samples; thereby realizing safe and efficient self-service sampling of asynchronous multi-threading. The present application can realize self-service, parallel multi-person, continuous multi-frequency sampling through asynchronous multi-threading, and the person to be tested can complete the sampling work at home by themselves, thereby reducing the risk of virus transmission.

The present invention relates to artificial intelligence understanding and environmental interaction, and in particular to an adaptive time aggregation network and method for online video visual relationship detection. The dynamic buffer memory used in the present invention to store historical video clips can fix the size of the storage content, and the storage content will not become larger and larger due to time. The step sampling strategy adopted by the present invention sets the sampling frequency by the strength of the correlation between the video frame and the current key frame, which can reduce the computational consumption required by the network while maintaining the accuracy of the detection results of the relationship between objects in the video in the time dimension. The adaptive spatiotemporal activation module and the attention-based knowledge state fusion module proposed in the present invention enable the invention to adaptively extract and fuse historical information and current status, and can detect the dynamic and static relationships of objects.

The invention discloses an online audio-video acquisition system and data analysis platform for clinical tests related to central venous catheterization. The online audio-video acquisition system comprises a clinical test announcement module, a test forecast module, an audio-video acquisition module, a data processing module, a key data generation module, a user management module, a user message module and user side software. The video and audio acquisition module and the video and audio storage module are matched for use, so that the test process can be recorded and played online in real time; through cooperative use of the data processing module and the key data generation module, test data can be marked and traced, and all key data needing to be collected are filled into a table; through cooperative use of the user management module, the user message leaving module and the user side software, online consulting, learning and communication of users can be realized, so that the test efficiency is effectively ensured, and questions of peers are prevented from being caused; and meanwhile, the advanced technology can be learned by the peer in real time, and the propagation of the medical technology is accelerated.

The application discloses a wireless charging motion mechanism, a rear handrail box and a car, which comprise a wireless charging assembly, a forward and reverse driving motor, a base, an X-direction transmission mechanism, a Z-direction overturning mechanism and a manual overturning mechanism; the forward and reverse driving motor is fixed at the end of the base, and the forward and reverse driving motor is in transmission connection with the end of the X-direction transmission mechanism; the front end of the X-direction transmission mechanism is pivotally connected with the rear end of the lower surface of the wireless charging assembly, the wireless charging assembly is in sliding connection with the base through the pivot of the X-direction transmission mechanism, and the wireless charging assembly can displace forward and backward along the length direction of the base; one end of the Z-direction overturning mechanism is hinged to the front end of the base, the other end of the Z-direction overturning mechanism is hinged to the middle part of the lower surface of the wireless charging assembly, and the Z-direction overturning mechanism and the X-direction transmission mechanism cooperate to enable the front end of the wireless charging assembly to overturn towards the Z direction; the application can adjust the inclination angle to carry out online video conference; the wireless charging mechanism can also manually rotate by maximum ± 90 DEG, and a mobile phone can be used in a horizontal screen mode.

The present invention provides a driver behavior monitoring method based on an online video understanding network, including: acquiring online video data; inputting the current frame image data into the feature extraction network in the online video understanding network to obtain the feature extraction result of the current frame image data; covering the feature extraction result of the current frame image data with a set part of the feature extraction results of historical frame image data to obtain the combined feature extraction result of the current frame image data; inputting the combined feature extraction result into the target prediction network in the online video understanding network to obtain the event monitoring task score of the current frame image data; and obtaining the driver behavior monitoring result according to the event monitoring task score. The present invention combines the features of each frame with the features of all historical frames without consuming too much computing resources, and performs target prediction based on the combined features, thereby realizing lightweight, high-efficiency, fast iteration, and good continuity of driver behavior monitoring.

1. Name of the product in this design: Wireless Car Box (A). 2. Purpose of this design: To enable functions such as mobile phone interconnection and online video through the original vehicle's in-vehicle channel. 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.

Systems and methods train, using training data, a prediction model by iteratively predicting a target variable value of an engagement event associated with an online video game application, identifying an error between a prediction and the target variable value, and modifying weights of the prediction model for multiple iterations. The training data includes information obtained from the online video game application and a partner computer application that stores resource data of real-world resources to a real-world location and also includes specific data such as a duration of gameplay, a quantity of instances of gameplay, and a quantity of resource transactions of the virtual resource. The trained prediction model is deployed and applied to user data of users to predict attributes of the engagement event that are most likely to cause users to engage with the online video game application within a predefined number of days.

1. The name of the design product: wireless car box (CP031-3). 2. The use of the design product: through the original vehicle channel, realize the function of mobile phone interconnection and online video, etc. 3. The design points of the design product: in shape. 4. The picture or photo that can best show the design points: perspective drawing.

Presented herein are stage user replacement techniques that can be performed to facilitate automatic stage user replacement for one or more stage users of an online video conference. In one example, a computer-implemented method is provided that may include providing, to a particular user interface of a particular participant, a management interface element that enables the particular participant to manage the at least one current stage user displayed in a stage display area for a communication session. Upon receiving at least one user interaction via the management interface element, the method may include automatically: replacing the at least one current stage user with one or more new stage users and synchronizing the stage display area of the particular user interface to be displayed on other stage display areas of other user interfaces of other participants of the session.