知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

18 records (took 0.017 seconds)

    REGULARIZING TARGETS IN MODEL DISTILLATION UTILIZING PAST STATE KNOWLEDGE TO IMPROVE TEACHER-STUDENT MACHINE LEARNING MODELS
    1.
    发明公开
    REGULARIZING TARGETS IN MODEL DISTILLATION UTILIZING PAST STATE KNOWLEDGE TO IMPROVE TEACHER-STUDENT MACHINE LEARNING MODELS 审中-公开

    公开(公告)号：US20240062057A1

    公开(公告)日：2024-02-22

    申请号：US17818506

    申请日：2022-08-09

    Applicant: Adobe Inc.

    Inventor： Surgan Jandial ,  Nikaash Puri ,  Balaji Krishnamurthy

    IPC: G06N3/08 ,  G06N3/04

    CPC classification number: G06N3/08 ,  G06N3/0454

    Abstract: This disclosure describes one or more implementations of systems, non-transitory computer-readable media, and methods that regularize learning targets for a student network by leveraging past state outputs of the student network with outputs of a teacher network to determine a retrospective knowledge distillation loss. For example, the disclosed systems utilize past outputs from a past state of a student network with outputs of a teacher network to compose student-regularized teacher outputs that regularize training targets by making the training targets similar to student outputs while preserving semantics from the teacher training targets. Additionally, the disclosed systems utilize the student-regularized teacher outputs with student outputs of the present states to generate retrospective knowledge distillation losses. Then, in one or more implementations, the disclosed systems compound the retrospective knowledge distillation losses with other losses of the student network outputs determined on the main training tasks to learn parameters of the student networks.

    Cloth warping using multi-scale patch adversarial loss
    2.
    发明授权
    Cloth warping using multi-scale patch adversarial loss 有权

    公开(公告)号：US11080817B2

    公开(公告)日：2021-08-03

    申请号：US16673574

    申请日：2019-11-04

    Applicant: Adobe Inc.

    Inventor： Kumar Ayush ,  Surgan Jandial ,  Mayur Hemani ,  Balaji Krishnamurthy ,  Ayush Chopra

    IPC: G06T3/00 ,  G06T11/60 ,  G06T7/40

    Abstract: Generating a synthesized image of a person wearing clothing is described. A two-dimensional reference image depicting a person wearing an article of clothing and a two-dimensional image of target clothing in which the person is to be depicted as wearing are received. To generate the synthesized image, a warped image of the target clothing is generated via a geometric matching module, which implements a machine learning model trained to recognize similarities between warped and non-warped clothing images using multi-scale patch adversarial loss. The multi-scale patch adversarial loss is determined by sampling patches of different sizes from corresponding locations of warped and non-warped clothing images. The synthesized image is generated on a per-person basis, such that the target clothing fits the particular body shape, pose, and unique characteristics of the person.

    Cloth Warping Using Multi-Scale Patch Adversarial Loss
    3.
    发明申请
    Cloth Warping Using Multi-Scale Patch Adversarial Loss 有权

    公开(公告)号：US20210133919A1

    公开(公告)日：2021-05-06

    申请号：US16673574

    申请日：2019-11-04

    Applicant: Adobe Inc.

    Inventor： Kumar Ayush ,  Surgan Jandial ,  Mayur Hemani ,  Balaji Krishnamurthy ,  Ayush Chopra

    IPC: G06T3/00 ,  G06T11/60 ,  G06T7/40

    Abstract: Generating a synthesized image of a person wearing clothing is described. A two-dimensional reference image depicting a person wearing an article of clothing and a two-dimensional image of target clothing in which the person is to be depicted as wearing are received. To generate the synthesized image, a warped image of the target clothing is generated via a geometric matching module, which implements a machine learning model trained to recognize similarities between warped and non-warped clothing images using multi-scale patch adversarial loss. The multi-scale patch adversarial loss is determined by sampling patches of different sizes from corresponding locations of warped and non-warped clothing images. The synthesized image is generated on a per-person basis, such that the target clothing fits the particular body shape, pose, and unique characteristics of the person.

    PERSONALIZED FORM ERROR CORRECTION PROPAGATION
    4.
    发明公开
    PERSONALIZED FORM ERROR CORRECTION PROPAGATION 审中-公开

    公开(公告)号：US20240362941A1

    公开(公告)日：2024-10-31

    申请号：US18140143

    申请日：2023-04-27

    Applicant: Adobe Inc.

    Inventor： Silky Singh ,  Surgan Jandial ,  Shripad Vilasrao Deshmukh ,  Milan Aggarwal ,  Mausoom Sarkar ,  Balaji Krishnamurthy ,  Arneh Jain ,  Abhinav Java

    IPC: G06V30/262 ,  G06V30/14 ,  G06V30/19 ,  G06V30/414

    CPC classification number: G06V30/274 ,  G06V30/1444 ,  G06V30/19147 ,  G06V30/414

    Abstract: A corrective noise system receives an electronic version of a fillable form generated by a segmentation network and receives a correction to a segmentation error in the electronic version of the fillable form. The corrective noise system is trained to generate noise that represents the correction and superimpose the noise on the fillable form. The corrective noise system is further trained to identify regions in a corpus of forms that are semantically similar to a region that was subject to the correction. The generated noise is propagated to the semantically similar regions in the corpus of forms and the noisy corpus of forms is provided as input to the segmentation network. The noise causes the segmentation network to accurately identify fillable regions in the corpus of forms and output a segmented version of the corpus of forms having improved fidelity without retraining or otherwise modifying the segmentation network.

    FORM STRUCTURE SIMILARITY DETECTION
    5.
    发明公开
    FORM STRUCTURE SIMILARITY DETECTION 审中-公开

    公开(公告)号：US20240330351A1

    公开(公告)日：2024-10-03

    申请号：US18190686

    申请日：2023-03-27

    Applicant: Adobe Inc.

    Inventor： Abhinav Java ,  Surgan Jandial ,  Shripad Vilasrao Deshmukh ,  Milan Aggarwal ,  Mausoom Sarkar ,  Balaji Krishnamurthy ,  Arneh Jain

    IPC: G06F16/383 ,  G06F16/332 ,  G06V30/19 ,  G06V30/412

    CPC classification number: G06F16/383 ,  G06F16/332 ,  G06V30/19147 ,  G06V30/412

    Abstract: Form structure similarity detection techniques are described. A content processing system, for instance, receives a query snippet that depicts a query form structure. The content processing system generates a query layout string that includes semantic indicators to represent the query form structure and generates candidate layout strings that represent form structures from a target document. The content processing system calculates similarity scores between the query layout string and the candidate layout strings. Based on the similarity scores, the content processing system generates a target snippet for display that depicts a form structure that is structurally similar to the query form structure. The content processing system is further operable to generate a training dataset that includes image pairs of snippets depicting form structures that are structurally similar. The content processing system utilizes the training dataset to train a machine learning model to perform form structure similarity matching.

    Model training with retrospective loss
    9.
    发明授权
    Model training with retrospective loss 有权

    公开(公告)号：US11797823B2

    公开(公告)日：2023-10-24

    申请号：US16793551

    申请日：2020-02-18

    Applicant: Adobe Inc.

    Inventor： Ayush Chopra ,  Balaji Krishnamurthy ,  Mausoom Sarkar ,  Surgan Jandial

    IPC: G06N3/04 ,  G06N3/084 ,  G06F18/214 ,  G06N3/047 ,  G06V10/764 ,  G06V10/82

    CPC classification number: G06N3/04 ,  G06F18/214 ,  G06N3/047 ,  G06N3/084 ,  G06V10/764 ,  G06V10/82

    Abstract: Generating a machine learning model that is trained using retrospective loss is described. A retrospective loss system receives an untrained machine learning model and a task for training the model. The retrospective loss system initially trains the model over warm-up iterations using task-specific loss that is determined based on a difference between predictions output by the model during training on input data and a ground truth dataset for the input data. Following the warm-up training iterations, the retrospective loss system continues to train the model using retrospective loss, which is model-agnostic and constrains the model such that a subsequently output prediction is more similar to the ground truth dataset than the previously output prediction. After determining that the model's outputs are within a threshold similarity to the ground truth dataset, the model is output with its current parameters as a trained model.

Patent Agency Ranking