公开(公告)号：US20220335638A1

公开(公告)日：2022-10-20

申请号：US17596794

申请日：2021-04-19

Applicant: Google LLC

Inventor： Abhishek Kar ,  Hossam Isack ,  Adarsh Prakash Murthy Kowdle ,  Aveek Purohit ,  Dmitry Medvedev

IPC: G06T7/593 ,  G06T7/285

Abstract: According to an aspect, a method for depth estimation includes receiving image data from a sensor system, generating, by a neural network, a first depth map based on the image data, where the first depth map has a first scale, obtaining depth estimates associated with the image data, and transforming the first depth map to a second depth map using the depth estimates, where the second depth map has a second scale.

公开(公告)号：US20250111477A1

公开(公告)日：2025-04-03

申请号：US18477219

申请日：2023-09-28

Applicant: GOOGLE LLC

Inventor： Sergio Orts Escolano ,  Zhiwen Fan ,  Di Qiu ,  Yinda Zhang ,  Daoye Wang ,  Erroll Wood ,  Abhimitra Meka ,  Hossam Isack ,  Paulo Fabiano Urnau Gotardo ,  Kripasindhu Sarkar ,  Thabo Beeler ,  Zhengyang Shen ,  Alexander Sahba Koumis

IPC: G06T5/50 ,  G06T7/194 ,  H04N5/272

Abstract: A method including capturing a first plurality of images that include a foreground object and a background, capturing a second plurality of images that include the background, generating an alpha matte based on the first plurality of images and the second plurality of images using a trained machine learned model trained using a loss function configured to cause the trained machine learned model to learn high-frequency details of the foreground object, generating a foreground object image based on the first plurality of images and the second plurality of images using the trained machine learned model, and synthesizing an image including the foreground object image and a second background scene using the alpha matte.

公开(公告)号：US11335023B2

公开(公告)日：2022-05-17

申请号：US15929811

申请日：2020-05-22

Applicant: Google LLC

Inventor： Sameh Khamis ,  Christian Haene ,  Hossam Isack ,  Cem Keskin ,  Sofien Bouaziz ,  Shahram Izadi

IPC: G06K9/00 ,  G06T7/73 ,  G06T3/40 ,  G06T11/00 ,  G06N3/08 ,  G06N3/04

Abstract: According to an aspect, a method for pose estimation using a convolutional neural network includes extracting features from an image, downsampling the features to a lower resolution, arranging the features into sets of features, where each set of features corresponds to a separate keypoint of a pose of a subject, updating, by at least one convolutional block, each set of features based on features of one or more neighboring keypoints using a kinematic structure, and predicting the pose of the subject using the updated sets of features.

公开(公告)号：US20210366146A1

公开(公告)日：2021-11-25

申请号：US15929811

申请日：2020-05-22

Applicant: Google LLC

Inventor： Sameh Khamis ,  Christian Haene ,  Hossam Isack ,  Cem Keskin ,  Sofien Bouaziz ,  Shahram Izadi

IPC: G06T7/73 ,  G06T3/40 ,  G06T11/00 ,  G06N3/04 ,  G06N3/08

Abstract: According to an aspect, a method for pose estimation using a convolutional neural network includes extracting features from an image, downsampling the features to a lower resolution, arranging the features into sets of features, where each set of features corresponds to a separate keypoint of a pose of a subject, updating, by at least one convolutional block, each set of features based on features of one or more neighboring keypoints using a kinematic structure, and predicting the pose of the subject using the updated sets of features.