中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

812 records (took 0.023 seconds)

    NEURAL NETWORK MODEL FOR AUDIO TRACK LABEL GENERATION
    261.
    发明公开
    NEURAL NETWORK MODEL FOR AUDIO TRACK LABEL GENERATION 审中-公开

    公开(公告)号：US20230386437A1

    公开(公告)日：2023-11-30

    申请号：US17804198

    申请日：2022-05-26

    Applicant: Lemon Inc.

    Inventor： Ju-Chiang Wang ,  Yun-Ning Hung ,  Jordan Smith

    IPC: G10H1/00 ,  G06N3/08

    CPC classification number: G10H1/0008 ,  G06N3/08 ,  G10H2210/056

    Abstract: System and methods directed to identifying music theory labels for audio tracks are described. More specifically, a first training set of audio portions may be generated from a plurality of audio tracks, segments within the plurality of audio tracks being labeled according to a plurality of music theory labels. A deep neural network model may then be trained using the first training set as an input, a first loss function for music theory label identifications of audio portions of the first training set, and a second loss function for segment boundary identifications within the audio portions of the first training set. In examples, the music theory label identifications and the segment boundary identifications are generated by the deep neural network model. A first audio track is received and segment boundary identifications and music theory labels for segments within the first audio track are generated using the deep neural network model.

    Procedural pattern generation for layered two-dimensional augmented reality effects
    262.
    发明授权
    Procedural pattern generation for layered two-dimensional augmented reality effects 有权

    公开(公告)号：US11830106B2

    公开(公告)日：2023-11-28

    申请号：US17531462

    申请日：2021-11-19

    Applicant: Lemon Inc.

    Inventor： Nathan Schager ,  Yixin Zhao

    IPC: G06T11/00 ,  G06T7/12

    CPC classification number: G06T11/00 ,  G06T7/12 ,  G06T2207/10016 ,  G06T2207/30196

    Abstract: Methods, systems and storage media for applying a pattern application effect to one or more frames of video are disclosed. Some examples may include: obtaining video data including one or more video frames, determining one or more segments in each of the one or more video frames, determining one or more object masks based on the one or more segments in each of the one or more video frames, combining, the one or more object masks into a single mask, obtaining pattern information, the pattern information representing one or more graphical effects to be applied to at least one layer of the one or more video frames, applying the pattern information to the single mask to generate masked pattern information and generating, by the computing device, a rendered video by adding the masked pattern information to the one or more video frames.

    APPROACH TO AUTOMATIC MUSIC REMIX BASED ON STYLE TEMPLATES
    265.
    发明公开
    APPROACH TO AUTOMATIC MUSIC REMIX BASED ON STYLE TEMPLATES 审中-公开

    公开(公告)号：US20230360619A1

    公开(公告)日：2023-11-09

    申请号：US17737282

    申请日：2022-05-05

    Applicant: Lemon Inc.

    Inventor： Bochen LI ,  Vibert THIO ,  Haonan CHEN ,  Xuefan HU ,  Jitong CHEN

    IPC: G10H1/00

    CPC classification number: G10H1/0008 ,  G10H1/0066 ,  G10H2210/131 ,  G10H2240/075 ,  G10H2240/131

    Abstract: In examples, a method for generating a remixed audio sample is provided. The method may include receiving an audio portion, obtaining metadata from the received audio portion, and analyzing the metadata and generating a symbolic music representation based on the analyzed metadata. In some examples, a selection of a style asset is received and applied to the symbolic music representation. Accordingly, a remixed audio portion may be rendered based on the stylized symbolic representation. That is, metadata associated with a song or song portion may be analyzed to identify a tempo, key, structure, chord, and/or progressions, etc., such that a remixed version of the song can be provided with customized instrumental arrangements and styles.

    DECENTRALIZED PROCEDURAL DIGITAL ASSET CREATION IN AUGMENTED REALITY APPLICATIONS
    266.
    发明公开
    DECENTRALIZED PROCEDURAL DIGITAL ASSET CREATION IN AUGMENTED REALITY APPLICATIONS 审中-公开

    公开(公告)号：US20230360280A1

    公开(公告)日：2023-11-09

    申请号：US17737569

    申请日：2022-05-05

    Applicant: Lemon Inc.

    Inventor： Nick NEWBERG ,  Peilin LI ,  Runjia TIAN ,  Nathanael SCHAGER ,  Gamze INANC

    IPC: G06T11/00

    CPC classification number: G06T11/00 ,  G06T2200/24

    Abstract: Example implementations include a method, apparatus and computer-readable medium for decentralized procedural digital asset creation, comprising receiving a first request to create a digital asset from an application executing an augmented reality effect on a computing device, wherein the first request includes an identifier associated with a user of the application. The implementations further include generating the digital asset and metadata of the digital asset, wherein the metadata includes information about characteristics of the digital asset and ownership of the digital asset by the user. Additionally, the implementations further include storing the metadata on a blockchain. Additionally, the implementations further include receiving a second request to access the digital asset from the application. Additionally, the implementations further include transmitting, to the application for rendering, the metadata stored on the blockchain in response to validating the identifier associated with the user in the second request.

    ATTRIBUTE AND RATING CO-EXTRACTION
    267.
    发明公开
    ATTRIBUTE AND RATING CO-EXTRACTION 审中-公开

    公开(公告)号：US20230342553A1

    公开(公告)日：2023-10-26

    申请号：US17727015

    申请日：2022-04-22

    Applicant: LEMON INC.

    Inventor： Hongyu XIONG ,  Han WANG ,  Yiqi FENG ,  Rui LI ,  Yuan GAO ,  Xiangyu ZENG ,  Qingyi LU ,  Bin LIU

    IPC: G06F40/279 ,  G06F40/30 ,  G06N3/04

    CPC classification number: G06F40/30 ,  G06F40/279 ,  G06N3/0454

    Abstract: Embodiments of the present disclosure relate to attribute and rating co-extraction. According to embodiments of the present disclosure, a method is proposed. The method comprises: determining, by a first sub-network of a model, a first feature representation based on a first token contained in a text, the first feature representation indicating semantic information of the first token in the text; determining, by a second sub-network of the model, first attribute information associated with the first token based on the first feature representation, the first attribute information indicating a first attribute involved in the text; and determining, by a third sub-network of the model, first rating information associated with the first token based on the first feature representation, the first rating information indicating a rating related to the first attribute.

    MODEL TRAINING BASED ON SYNTHETIC DATA
    269.
    发明公开
    MODEL TRAINING BASED ON SYNTHETIC DATA 审中-公开

    公开(公告)号：US20230334834A1

    公开(公告)日：2023-10-19

    申请号：US18338056

    申请日：2023-06-20

    Applicant: Lemon Inc. ,  Beijing Youzhuju Network Technology Co., Ltd.

    Inventor： Song BAI ,  Ruifei HE ,  Shuyang SUN ,  Xin YU ,  Chuhui XUE ,  Wenqing ZHANG ,  Xiaojuan QI

    IPC: G06V10/774 ,  G06V10/764 ,  G06T11/60

    CPC classification number: G06V10/774 ,  G06V10/764 ,  G06T11/60

    Abstract: Embodiments of the present disclosure relate to model training based on synthetic data. According to example embodiments of the present disclosure, synthetic images are generated by providing respective text prompts into a text-to-image generation model. Respective training labels associated with the synthetic images are also generated based on the used text prompts. A target model, which is configured to perform an image classification task, is trained based at least in part on the synthetic images and the associated training labels. Through this solution, a large scale of synthetic images can be automatically obtained and applicable for training a model for image classification, to improve the model performance with data-scare setting or in the case of model pre-training where the training data amount matters.

Patent Agency Ranking