19 results about "Consecutive frame" patented technology

The invention provides a continuous frame parameter adjustment method, electronic equipment and a system in on-board diagnostic communication. The method comprises the following steps: determining the priority of UDS service; determining a reference value of the first parameter according to the priority of the UDS service; determining a predicted value of a network load at a future moment through a network load prediction model based on the correlation characteristics of the vehicle at the current moment; and correcting the reference value of the first parameter according to the predicted value of the network load at the future moment and the true value of the network load at the current moment to obtain a corrected value of the first parameter. Through the technical scheme of the invention, the continuous frame parameter can adapt to the dynamically fluctuating network load, the loss probability of the continuous frame is reduced, and the stable and effective transmission of the continuous frame is ensured.

Approaches are disclosed for enhancing frame rate and visual smoothness in real-time video streams through multi-frame interpolation. A classification neural network analyzes two sequential frames, outputting confidence scores that indicate the reliability of motion data for each pixel. These scores determine whether a pixel's motion is accurately described by motion vectors or should be treated as static. The classification results are reused to generate intermediate frames by warping the original frames based on the motion characteristics. Blending weights are calculated by combining warped motion vector confidence values with static values, and a second neural network refines the alignment and blending of candidate frames. This second network predicts intermediate flows and generates new blending weights, which are used to warp and blend the candidate frames, ultimately producing a final interpolated frame that enhances visual smoothness and consistency in the video stream.

A method for compressing a video stream includes retrieving a plurality of frames corresponding to the video stream. For each of two or more sequential frames of the plurality of frames of the video stream, the method includes extracting Key Point Descriptors (KPDs) for the respective frame and processing the respective frame using Principle Component Analysis (PCA) followed by vector quantization, resulting in a quantized explained variance matrix for the respective frame. The quantized explained variance matrix for the respective frame is stored. The KPDs for the respective frame are stored.

This invention provides a system and method for providing video semantic segmentation. Semantic segmentation is performed on a first frame in a sequence of video frames to obtain at least one first semantic feature of the first frame. Semantic segmentation is performed on a second frame in the sequence to obtain at least one second semantic feature of the second frame, wherein the second frame follows the first frame. Semantic segmentation is performed on a third frame in the sequence to obtain at least one third semantic feature, wherein the third frame follows and is also following the first frame by a first predetermined number of consecutive frames. At least one first semantic feature, at least one second semantic feature, and at least one third semantic feature are combined to form at least one fourth semantic feature of the second frame.

The application relates to a video translation method and device, which comprises the following steps: inputting image features and coordinate features of original data into a fusion input generator; calculating a three-frame loss according to continuous frames of a video of the input generator, and optimizing the generator according to the three-frame loss; and performing video translation based on the optimized generator, so as to improve the visual effect of the converted video.

Disclosed are an encoder for encoding a video sequence having successive frames into a bitstream, a method for encoding a video sequence having successive frames into a bitstream, a decoder for decoding the bitstream to reconstruct the video sequence having successive frames, and a method for decoding the bitstream to reconstruct the video sequence having successive frames.

To generate communication specifications of a device whose communication specifications are unknown.SOLUTION: A communication specification generation device (101) includes a frame extraction unit (11) that extracts a frame having a specific frame length from communication log data generated by a target device (103), a header identification unit (12) that sequentially compares values in a range determined by a standard header length from a head of the frame between two consecutive frames to identify a header of the frame, and a combination identification unit (13) that extracts codes of a command and a response from the header on the basis of position information and size information about the respective codes of the command and response frames and identifies a combination of the codes of the command and the response according to a communication state of the command and response frames in the communication log data.SELECTED DRAWING: Figure 1

The invention discloses a detection method and device, electronic equipment, a storage medium and a computer program product. The method comprises: determining M segments of frame sequences from a video to be detected, M being an integer greater than or equal to 1, each segment of frame sequence in the M segments of frame sequences including N continuous frames, the N continuous frames included in each segment of frame sequence being completely different, and N being an integer greater than 1; for each section of frame sequence in the M sections of frame sequences, performing mask processing on N continuous frames contained in the frame sequence in the to-be-detected video to obtain a mask video; by using data contained in the mask video, sequentially regenerating N continuous frames to be masked according to a sequence from two ends to the middle; detecting the frame sequence by using N continuous frames contained in the frame sequence and the regenerated N continuous frames to obtain a detection result corresponding to the frame sequence; and determining whether the to-be-detected video belongs to a forged video or not by using all the detection results corresponding to the M segments of frame sequences to obtain a first detection result.

The invention relates to a flying dust detection method and device, electronic equipment and a readable storage medium. The method comprises the following steps: respectively acquiring a radar point cloud and a depth map of a current frame through a radar and a depth camera; wherein the radar point cloud and the depth map of the same frame correspond to each other and form a parameter group; updating a frame queue of the current frame based on the parameter group, the frame queue being formed by a plurality of parameter groups corresponding to continuous frames, and the frame queue comprising the parameter group of the current frame; and determining a fusion score of the parameter group of the current frame, and determining a target detection result of the current frame based on the fusion scores of all the parameter groups in the frame queue. By adopting the method, the dust raising environment can be accurately detected.

This disclosure relates to multi-frame interpolation for real-time video processing using deep neural networks. A method is disclosed to improve frame rate and visual smoothness in real-time video streams through multi-frame interpolation. A classification neural network analyzes two consecutive frames and outputs confidence scores indicating the reliability of motion data for each pixel. These scores determine whether the motion of a pixel is accurately described by a motion vector or should be considered static. The classification results are reused to generate intermediate frames by warping the original frames based on motion features. Blending weights are calculated by combining the warped motion vector confidence values ​​with static values, and a second neural network refines the alignment and blending of candidate frames. This second network predicts the intermediate stream and generates new blending weights, which are used to warp and blend candidate frames, ultimately producing final interpolated frames that enhance the visual smoothness and consistency of the video stream.

Techniques for frame-based noise reduction and other signal processing techniques are provided that are able to reduce latency below the frame length. This may be accomplished by using an asymmetric trapezoidal (or quasi-trapezoidal) filter. Instead of cross-fading between successive frames taking up half the frame length, the cross-fading between successive frames is reduced to 10-25% of the frame length, for example. This allows the latency to be reduced to 60-75% of the frame length (plus the signal processing time) instead of the full frame length (plus the signal processing time) in prior techniques.

The invention discloses a bird identification method, which comprises the following steps: according to a bird detection result of a to-be-identified data frame, a bird identification result and continuous frame information of at least one continuous frame of the to-be-identified data frame, at least one effective frame is obtained, and the effective frame is used for representing a data frame which detects a bird target and is stable; under the condition that the frame number of the obtained effective frames is larger than the set frame number, the obtained effective frames are searched according to the set frame number, the bird recognition results corresponding to the currently searched effective frames are counted, the statistical result of the number of times of all the bird recognition results is obtained, and the bird recognition results are used for representing the types of birds; and selecting a bird recognition result with the maximum number of times from the statistical results, and taking the selected bird recognition result as a current recognition result under the condition that the maximum number of times of the selected bird recognition result is greater than a set first number of times threshold value. According to the invention, false detection and false identification can be prevented, and the bird identification precision is improved.

Systems and methods are provided for one or more processors; and memory storing instructions that, when executed by the one or more processors, cause the system to perform: receiving successive frames of sensor data, the successive frames comprising a first frame and a second frame; determining transformations, in sensor coordinates, between coordinates of corresponding elements in the successive frames; determining a mapping between the transformations in sensor coordinates and transformations in geospatial coordinates of the corresponding elements in the successive frames; and determining second geospatial coordinates of the corresponding elements of a third frame based on: a transformation between the second frame and the third frame, and the mapping.

Systems and methods for universal stage identification for intra-operative and post-operative applications are provided. The system receives a video stream that captures a program of the robotic medical system over a time interval. A system generates a set of consecutive frames from a video stream. The first set of consecutive frames may include a first temporal resolution and the second set of consecutive frames may include a second temporal resolution. The system determines, on a transient basis, a stage of the program within a time interval via a set of consecutive frames input into a first model trained with machine learning. The system inputs stages of the program into the second model to generate stage segments within the time interval. The second model may be trained with machine learning based on historical workflows. The system provides actions based on the metrics of the stage segments.

Disclosed is an encoder for encoding a video sequence comprising consecutive frames into a bit stream, a method for encoding a video sequence comprising consecutive frames into a bit stream, a decoder for decoding a bit stream in order to reconstruct a video sequence comprising consecutive frames and a method for decoding a bit stream in order to reconstruct a video sequence comprising consecutive frames.

A method includes determining an operating mode of a first terminal of the one or more terminals; selecting, based on the operating mode, a first frame configuration from a plurality of frame configurations for the first terminal, each of the plurality of frame configurations configuring one or more of: an offset parameter determining the start of a frame, a Channel Occupancy Time “COT” duration in a frame, an idle duration in a frame, a frame duration, a frame period and a gap duration between two subsequent frames; and communicating between the first terminal and the base station and via the first frequency band, using contention-based access and based on the first frame configuration.

This invention belongs to the field of video coding technology, specifically relating to a method for fast coding unit partitioning in 360-degree video based on temporal correlation. The method includes: acquiring and preprocessing cross-frame correlation data, including the current frame CU, the previous frame CU, the luminance difference map between the current and previous frames, the partitioning result of the previous frame, and the partitioning result label; processing the luminance difference map using a similarity network to obtain the similarity between the current and previous frames; selecting, based on the similarity, to directly obtain the coding unit partitioning result from the previous frame's partitioning result or selecting to process the cross-frame correlation data using a multi-level adaptive differentiation prediction network to obtain the coding unit partitioning prediction probability; and determining the partitioning mode based on the coding unit partitioning prediction probability. This invention can operate stably in complex video scenes and under different dynamic scenarios, accurately identify the structural similarity between consecutive frames, and directly reuse the partitioning of the previous frame in highly similar regions, avoiding redundant calculations and improving coding efficiency and stability.

The invention relates to a method for realizing GIF (Graphics Interchange Format) adaptive steganography based on a spatio-temporal policy network and cross-frame optimization, which comprises the following steps of: inputting continuous multi-frame images serving as frame fragments into a pre-trained spatio-temporal policy network to generate policy information; determining the embedding capacity of secret information for the target image in the frame segment through a cross-frame optimization module; based on the determined embedding capacity and embedding cost distribution contained in the strategy information, embedding secret information into the target image; and the steganography GIF is analyzed, and the secret information is recovered. According to the method and the device for realizing the GIF adaptive steganography based on the space-time strategy network and the cross-frame optimization, the processor and the computer readable storage medium thereof, the space-time characteristics of the GIF are automatically learned by utilizing the strategy network obtained by deep reinforcement learning training, so that an optimal embedding strategy and a cross-frame load distribution scheme are generated, and the self-adaptive steganography of the GIF is realized. Therefore, the high embedding capacity is ensured, and meanwhile, the anti-detection safety of steganography is remarkably improved.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for compressing video data. In one aspect, a method comprises: receiving a video sequence of frames; generating, using a flow prediction network, an optical flow between two sequential frames, wherein the two sequential frames comprise a first frame and a second frame that is subsequent the first frame; generating from the optical flow, using a first autoencoder neural network: a predicted optical flow between the first frame and the second frame; and warping a reconstruction of the first frame according to the predicted optical flow and subsequently applying a blurring operation to obtain an initial predicted reconstruction of the second frame.