-
公开(公告)号：US20230401813A1
公开(公告)日：2023-12-14
申请号：US18035665
申请日：2021-11-08
Applicant: CANON U.S.A., INC.
Inventor： Xiwu Cao
CPC classification number: G06V10/431 , G06V10/25 , G06V10/32 , G06V10/993 , G06V20/70
Abstract: An image processing apparatus and method are provided which obtains an image captured by an image capture device and stored in a memory, extracts one or more regions of interest in the obtained image, normalizes the extracted one or more regions of interest to be at a same scale or some predefined scales, extract the frequency information of regions of interest, determines a sharpness of the obtained image by aggregating the frequency information in each of the one or more extracted regions of interest, and labels the obtained image with the determined sharpness score.
-
Patent Agency Ranking
公开(公告)号：US20250069323A1
公开(公告)日：2025-02-27
申请号：US18724538
申请日：2022-12-29
Applicant: CANON U.S.A., INC.
Inventor： Jason Mack Williams , Ryuhei Konno , Jonathan Forr Lorentz , Bradley Scott Denney , Xiwu Cao , Peng Sun , Quentin Dietz , Jeanette Yang Paek
Abstract: An information processing apparatus and method for relighting captured images in a virtual reality environment is provided. Righting processing performed by the apparatus includes acquiring an input image, acquiring a target image, determining at least one shared reference region for the acquired input image and at least one shared reference region for the acquired target image, determining a transform matrix based on the color space of the shared region, applying the transform matrix to at least a portion of the acquired input image, outputting the transformed portion of the acquired input image, and displaying the outputted transformed portion of the acquired input image.
-
公开(公告)号：US20250063142A1
公开(公告)日：2025-02-20
申请号：US18724104
申请日：2022-12-29
Applicant: CANON U.S.A., INC.
Inventor： Jason Mack Williams , Ryuhei Konno , Jonathan Forr Lorentz , Bradley Scott Denney , Xiwu Cao , Peng Sun , Quentin Dietz , Jeanette Yang Paek
Abstract: A system for immersive virtual reality communication that includes a first capture device configured to capture a stream of images of a first user, a first network configured to transmit the captured stream of images of the first user, a second network configured to receive data based at least in part on the captured stream of images of the first user and a first virtual reality device used by a second user, wherein the first virtual reality device is configured to render a virtual environment and to produce a rendition of the first user based at least in part on the data based at least in part on the stream of images of the first user produced by the first capture device.
-
公开(公告)号：US20250124650A1
公开(公告)日：2025-04-17
申请号：US18695666
申请日：2022-09-29
Applicant: CANON U.S.A., INC.
Inventor： Xiwu Cao
Abstract: A server and method is provided for removing an apparatus that occludes a portion of a face in a video stream and receives captured video data of a user wearing the apparatus that occludes the portion of the face of the user, obtains facial landmarks representing the entire face of the user including the occluded portion and non-occluded portion of the face of the user, provides one or more types of reference images of the user with the obtained facial landmarks to a trained machine learning model to remove the apparatus from the received captured video data, generates three dimensional data of the user including a full face image using the trained machine learning model and causes the generated three dimensional data of the user to be displayed on a display of the apparatus that occludes the portion of the face of the user.
-
公开(公告)号：US20230410478A1
公开(公告)日：2023-12-21
申请号：US18035635
申请日：2021-11-08
Applicant: CANON U.S.A., INC.
Inventor： Xiwu Cao
IPC: G06V10/766 , G06V10/776 , G06V10/764
CPC classification number: G06V10/766 , G06V10/776 , G06V10/764
Abstract: A processing apparatus is provided that is configured to perform operations including obtaining a plurality of images having been evaluated by different sources such that each source has classified each of the plurality of image as being a member of one of a predefined class, generating a distribution array identifying a number of times each image of the plurality of images has been classified into each of the predefined classes, generating, for each predefined class, a loss function based on the ratio of a number of images in other classes of the predefined classes to a number of images to this predefined classes, providing the generated loss function for each predefined class as evaluation parameters to a model, and using the generated loss function to determine that the model classifies raw image data as being a member of one of the predefined classes according to a predetermined accuracy threshold.
-
6.发明申请公开(公告)号：US20200372383A1
公开(公告)日：2020-11-26
申请号：US16422799
申请日：2019-05-24
Inventor： Xiwu Cao , Samuel Schrader
Abstract: Applying a local-adapted minority oversampling strategy technique to an imbalanced dataset including positive samples belonging to a minority class and negative samples belonging to a majority class, the minority class being less prevalent than the majority class, wherein each sample from the dataset includes a plurality of features. The local-adapted minority oversampling strategy includes determining a local imbalance for each positive sample from the minority class corresponding to a number of other positive samples and/or negative samples within a neighborhood of each positive sample. The local-adapted minority oversampling strategy also includes calculating a local-adapted oversampling ratio based on the local imbalance estimated for each positive sample and replicating each positive sample using the local-adapted oversampling ratio to generate a new dataset.
-
-
-
-
-
Search Results
6
records
(took 0.01 seconds)
