中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

12 records (took 0.015 seconds)

    FLOW-AGNOSTIC NEURAL VIDEO COMPRESSION
    1.
    发明公开
    FLOW-AGNOSTIC NEURAL VIDEO COMPRESSION 审中-公开

    公开(公告)号：US20230169694A1

    公开(公告)日：2023-06-01

    申请号：US17975471

    申请日：2022-10-27

    Applicant: QUALCOMM Incorporated

    Inventor： Hoang Cong Minh LE ,  Reza POURREZA ,  Yang YANG ,  Yinhao ZHU ,  Amir SAID ,  Taco Sebastiaan COHEN

    IPC: G06T9/00

    CPC classification number: G06T9/002

    Abstract: A processor-implemented method for video compression using an artificial neural network (ANN) includes receiving a video via the ANN. The ANN extracts a first set of features of a current frame of the video and a second set of features of a reference frame of the video. The ANN determines an estimate of correlation features between the first set of features of the current frame and the second set of features of the reference frame. The estimate of the correlation features are encoded and transmitted to a receiver.

    MULTI-SCALE OPTICAL FLOW FOR LEARNED VIDEO COMPRESSION
    3.
    发明申请
    MULTI-SCALE OPTICAL FLOW FOR LEARNED VIDEO COMPRESSION 有权

    公开(公告)号：US20220303568A1

    公开(公告)日：2022-09-22

    申请号：US17207244

    申请日：2021-03-19

    Applicant: QUALCOMM Incorporated

    Inventor： Reza POURREZA ,  Amir SAID ,  Yang YANG ,  Yinhao ZHU ,  Taco Sebastiaan COHEN

    IPC: H04N19/51 ,  H04N19/172 ,  H04N19/137 ,  H04N19/107 ,  H04N19/593 ,  G06N3/08

    Abstract: Systems and techniques are described for encoding and/or decoding data based on motion estimation that applies variable-scale warping. An encoding device can receive an input frame and a reference frame that depict a scene at different times. The encoding device can generate an optical flow identifying movements in the scene between the two frames. The encoding device can generate a weight map identifying how finely or coarsely the reference frame can be warped for input frame prediction. The encoding device can generate encoded video data based on the optical flow and the weight map. A decoding device can generate a reconstructed optical flow and a reconstructed weight map from the encoded data. A decoding device can generate a prediction frame by warping the reference frame based on the reconstructed optical flow and the reconstructed weight map. The decoding device can generate a reconstructed input frame based on the prediction frame.

    PROGRESSIVE DATA COMPRESSION USING ARTIFICIAL NEURAL NETWORKS
    5.
    发明申请
    PROGRESSIVE DATA COMPRESSION USING ARTIFICIAL NEURAL NETWORKS 有权

    公开(公告)号：US20220237740A1

    公开(公告)日：2022-07-28

    申请号：US17648808

    申请日：2022-01-24

    Applicant: QUALCOMM Incorporated

    Inventor： Yadong LU ,  Yang YANG ,  Yinhao ZHU ,  Amir SAID ,  Taco Sebastiaan COHEN

    IPC: G06T3/40 ,  G06T9/00

    Abstract: Certain aspects of the present disclosure provide techniques for compressing content using a neural network. An example method generally includes receiving content for compression. The content is encoded into a first latent code space through an encoder implemented by an artificial neural network trained to generate a latent space representation of the content. A first compressed version of the encoded content is generated using a first quantization bin size of a series of quantization bin sizes. A refined compressed version of the encoded content is generated by scaling the first compressed version of the encoded content into one or more second quantization bin sizes smaller than the first quantization bin size, conditioned at least on a value of the first compressed version of the encoded content. The refined compressed version of the encoded content is output for transmission.

    MEMORY REDUCTION FOR NON-SEPARABLE TRANSFORMS
    8.
    发明申请
    MEMORY REDUCTION FOR NON-SEPARABLE TRANSFORMS 审中-公开

    公开(公告)号：US20190007705A1

    公开(公告)日：2019-01-03

    申请号：US16020511

    申请日：2018-06-27

    Applicant: QUALCOMM Incorporated

    Inventor： Xin ZHAO ,  Vadim SEREGIN ,  Amir SAID ,  Marta KARCZEWICZ

    IPC: H04N19/61 ,  H04N19/176 ,  H04N19/50 ,  H04N19/18

    Abstract: Techniques are described in which a decoder is configured to receive an input data block and apply an inverse non-separable transform to at least part of the input data block to generate an inverse non-separable transform output coefficient block. The applying the inverse non-separable transform comprises assigning a window, assigning a weight for each position inside the assigned window, and determining the inverse non-separable transform output coefficient block based on the assigned weights. The decoder is further configured to forming a decoded video block based on the determined inverse non-separable transform output coefficient block, wherein forming the decoded video block comprises summing the residual video block with one or more predictive blocks.

Patent Agency Ranking