公开(公告)号：WO2022051474A2

公开(公告)日：2022-03-10

申请号：PCT/US2021/048837

申请日：2021-09-02

Applicant: CURIOUSER PRODUCTS INC.

Inventor： PUTNAM, Brynn ,  D'AMBROSIO-CORRELL, Kristie

IPC: A63B24/00 ,  G06K9/00 ,  A61B5/11 ,  A61B5/00 ,  G09B19/00 ,  A61B5/0205 ,  A47G1/02 ,  A61B5/0022 ,  A61B5/024 ,  A61B5/0816 ,  A61B5/1128 ,  A61B5/222 ,  A61B5/486 ,  A61B5/6891 ,  A61B5/7267 ,  A61B5/7275 ,  A61B5/743 ,  A61B5/744 ,  A63B2024/0009 ,  A63B2024/0065 ,  A63B2071/063 ,  A63B2071/0638 ,  A63B2071/0683 ,  A63B2220/12 ,  A63B2220/40 ,  A63B2220/75 ,  A63B2220/805 ,  A63B2220/806 ,  A63B2225/12 ,  A63B2230/06 ,  A63B2230/50 ,  A63B24/0006 ,  A63B24/0062 ,  A63B71/0622 ,  G06F3/011 ,  G06V40/103 ,  G06V40/23 ,  G09B19/0038

Abstract: A processor-implemented method includes receiving a request that specifies a workout performed at a first time, a list of users, a second time after the first time, an overlay to be displayed during a rebroadcast associated with the request, and a skill level of the workout. The request is compared to calendar data, and a session acknowledgment message is sent to the compute device of the first user based on the comparison. An invitation message is sent to compute devices of a second user and a third user, identifying the second time, and invitation responses are received from the second user and the third user. In response to the invitation responses, a video of the workout is rebroadcast at the second time, to a smart mirror of the first user, a smart mirror of the second user and a smart mirror of the third user.

公开(公告)号：WO2022010816A1

公开(公告)日：2022-01-13

申请号：PCT/US2021/040400

申请日：2021-07-05

Applicant: ECTO, INC.

Inventor： KOZACHENOK, Dmitry ,  TORNG, Allen

IPC: G06K9/00 ,  G06V20/00 ,  G06V40/103

Abstract: A method of surface splash scoring includes receiving, at an electronic device, a set of camera frames corresponding to images of a water surface. The electronic device processes the set of camera frames with a trained machine learning model to generate one or more quantifications associated with fish activity proximate the water surface. In some embodiments, a surface splash score is computed that represents an appetite level anticipated to be exhibited for a first time period. Subsequently, the electronic device generates an output indicative of the surface splash score.

公开(公告)号：WO2021245156A1

公开(公告)日：2021-12-09

申请号：PCT/EP2021/064833

申请日：2021-06-02

Applicant: ROBERT BOSCH GMBH

Inventor： MEHNERT, Jens Eric Markus

IPC: G06K9/00 ,  G06K9/46 ,  G06K9/52 ,  G06K9/62 ,  G06N3/04 ,  H04L12/40 ,  G06K9/627 ,  G06N3/0454 ,  G06N3/088 ,  G06N5/003 ,  G06V10/431 ,  G06V10/454 ,  G06V10/52 ,  G06V10/82 ,  G06V20/58 ,  G06V20/582 ,  G06V40/103 ,  H04L12/40032

Abstract: Vorrichtung (1) für die Klassifikation von Bilddaten (2), umfassend • eine trainierbare Vorverarbeitungseinheit (11, 11a-11d), welche dazu ausgebildet ist, anhand der Bilddaten (2) mindestens eine Vorgabe (3) dahingehend, inwieweit der Detaillierungsgrad der Bilddaten (2) zu reduzieren ist, aus einem trainierten Zusammenhang abzurufen (111) und den Detaillierungsgrad der Bilddaten (2) entsprechend dieser Vorgabe (3) zu reduzieren (112); sowie • einen trainierbaren Klassifikator (12), der dazu ausgebildet ist, die detailreduzierten Bilddaten (4) auf eine Zuordnung (5) zu einer oder mehreren Klassen einer vorgegebenen Klassifikation abzubilden. Verfahren (100) zum Trainieren der Vorrichtung (1), wobei Parameter (11*), die das Verhalten der trainierbaren Vorverarbeitungseinheit (11, 11a-11d) charakterisieren, werden auf die Ziele optimiert (130), dass die Vorrichtung (1) Lern-Bilddaten (2a) auf Lern-Zuordnungen (5a) abbildet und zugleich die Reduzierung des Detaillierungsgrades, die die Vorverarbeitungseinheit (11, 11a-11d) an den Lern-Bilddaten (2a) vornimmt, im Mittel einer Vorgabe (3a) entspricht.

公开(公告)号：WO2023274832A1

公开(公告)日：2023-01-05

申请号：PCT/EP2022/067154

申请日：2022-06-23

Applicant: FOTONATION LIMITED

Inventor： DINU, Dragos ,  FULOP, Szabolcs ,  ENE, Radu

IPC: G06V20/59 ,  G06V40/10 ,  G06V10/82 ,  B60W2420/40 ,  B60W2420/42 ,  B60W2540/221 ,  B60W2540/223 ,  B60W2540/229 ,  B60W40/08 ,  G06F18/21 ,  G06F18/24 ,  G06N3/045 ,  G06T2207/10048 ,  G06T2207/20084 ,  G06T2207/30196 ,  G06T2207/30268 ,  G06T3/40 ,  G06T7/70 ,  G06V10/147 ,  G06V10/22 ,  G06V20/597 ,  G06V40/103 ,  G06V40/107 ,  G06V40/161 ,  G06V40/28

Abstract: A method for monitoring occupants of a vehicle comprises identifying a respective body region for one or more occupants of the vehicle within an obtained image; identifying within the body regions, skeletal points including points on an arm of a body; identifying one or more hand regions; and determining a hand region to be either a left or a right hand of an occupant in accordance with its spatial relationship with identified skeletal points of the body region of an occupant. The left or right hand region for the occupant are provided to a pair of classifiers to provide an activity classification for the occupant, a first classifier being trained with images of hands of occupants in states where objects involved are not visible and a second classifier being trained with images of occupants in the states where the objects are visible in at least one hand region.

公开(公告)号：WO2022237688A1

公开(公告)日：2022-11-17

申请号：PCT/CN2022/091484

申请日：2022-05-07

Applicant: 影石创新科技股份有限公司

Inventor： 贾配洋 ,  侯俊

IPC: G06K9/00 ,  G06N3/045 ,  G06N3/048 ,  G06N3/08 ,  G06T3/4007 ,  G06V10/462 ,  G06V40/103

Abstract: 本申请涉及一种姿态估计方法、装置、计算机设备和存储介质。所述方法包括：获取待进行姿态估计的目标图像；所述目标图像中包括待处理的目标对象；基于所述目标图像进行特征提取，获取第一提取特征；通过图像特征扩张网络对所述第一提取特征进行特征扩张，得到扩张图像特征；对所述扩张图像特征进行特征提取，得到第二提取特征；通过图像特征压缩网络对所述第二提取特征进行特征压缩，得到压缩图像特征；基于所述压缩图像特征确定所述目标图像中的所述目标对象对应的关键点位置信息，基于所述关键点位置信息对所述目标对象进行姿态估计。本方法能够提高姿态估计的效率。

公开(公告)号：WO2022208440A1

公开(公告)日：2022-10-06

申请号：PCT/IB2022/053034

申请日：2022-03-31

Applicant: SONY GROUP CORPORATION ,  SONY CORPORATION OF AMERICA

Inventor： ZHANG, Qing ,  XIAO, Hanyuan

IPC: G06V40/10 ,  G06V10/82 ,  G06T15/08 ,  G06T19/20 ,  G06T13/40 ,  G06T17/00 ,  G06T2207/20084 ,  G06T2207/30196 ,  G06T2210/16 ,  G06T7/55 ,  G06T7/70 ,  G06V40/103

Abstract: A neural human performance capture framework (MVS-PERF) captures the skeleton,body shape and clothes displacement, and appearance of a person from a set of calibratedmultiview images. It addresses the ambiguity of predicting the absolute position in monocularhuman mesh recovery, and bridges the volumetric representation from NeRF toanimation-friendly performance capture. MVS-PERF includes three modules to extract featuremaps from multiview images and fuse them to a feature volume, regress the feature volume to anaked human parameters vector, generating an SMPL-X skin-tight body mesh with a skeletalpose, body shape, and expression, and leverage a neural radiance field and a deformation field toinfer the clothes as the displacement on the naked body using differentiable rendering. Clothedbody mesh is obtained by adding the interpolated displacement vectors to the SMPL-X skin-tightbody mesh vertices. The obtained radiance field is used for free-view volumetric rendering of theinput subject.

公开(公告)号：WO2023272993A1

公开(公告)日：2023-01-05

申请号：PCT/CN2021/121897

申请日：2021-09-29

Applicant: 苏州浪潮智能科技有限公司

Inventor： 范宝余 ,  王立

IPC: G06K9/62 ,  G06F18/214 ,  G06F18/22 ,  G06F18/241 ,  G06N3/045 ,  G06N3/08 ,  G06N3/084 ,  G06V10/40 ,  G06V40/103

Abstract: 本申请公开了一种图片识别方法、装置、设备及可读存储介质，该方法包括：获取待识别的目标图片；将目标图片输入至训练好的特征提取模型进行特征提取，得到图像特征；利用图像特征，对目标图片进行识别，得到识别结果。在本申请中，在无需增加特征提取模型的参数量和计算量的情况下，通过引出同构分支，特征相互挖掘即知识协同辅助训练，使得特征提取模型具有较佳的特征提取性能，可基于该特征提取模型所提取的图像特征完成更加准确的图片识别。

公开(公告)号：WO2022268537A1

公开(公告)日：2022-12-29

申请号：PCT/EP2022/065906

申请日：2022-06-10

Applicant: VALEO COMFORT AND DRIVING ASSISTANCE

Inventor： TAMAM, Mahmoud ,  QUESSADA, Audrey ,  RINGELSTEIN, Nicolas ,  ELKHATEEB, Mahmoud

IPC: G06V10/82 ,  G06V20/59 ,  G06V40/10 ,  G06V20/593 ,  G06V40/103

Abstract: The invention relates to a method (1) for defining the clothes type (t1) of an occupant (o) of a vehicle (2), wherein said method (1) comprising: - acquiring (E1) at least one video stream (Vs) of the cabin (20) of said vehicle (2), - performing (E2) skin segmentation from images (I) of said at least one video stream (Vs) to classify each image pixels (p) into skin (s0) versus non skin (s1) as human pixels (p0, p1) or into background pixels (p2), - detecting (E3) from said images (I) of said video stream (Vs) body key points (K;) per occupant (o) within said cabin (20), - determining (E4) from said body key points (K;) body parts (B) of each occupant (o) within said cabin (20), a body part (B) being a connection between at least two body key points (Kt), - for each image (I) of the video stream (Vs), calculating (E5) skin (s0) versus non skin (s1) percentage levels (L) over body parts (B) of each occupant (o), - for each occupant (o) of the vehicle (2), defining (E6) clothes type (t1) based on said skin (sO) versus non skin (s1) percentage levels (L) of each body part (B) of said occupant (o).

公开(公告)号：WO2022212183A1

公开(公告)日：2022-10-06

申请号：PCT/US2022/021825

申请日：2022-03-24

Applicant: SONY INTERACTIVE ENTERTAINMENT LLC

Inventor： ZHANG, Xinyu

IPC: A63F13/60 ,  G06T13/40 ,  G06T13/80 ,  G06T15/04 ,  G06V40/103 ,  G06V40/176 ,  G06V40/23

Abstract: Methods and systems for re-mastering animation files used in a video game includes identifying a rig used for representing a virtual character of the video game. Virtual markers are applied to the rig to generate a modified rig. Animation files for the virtual character are executed using the modified rig and a virtual camera is activated to capture images of the animation of the modified rig. Images of the modified rig are used to define performance data. The performance data is applied to a new rig of the virtual character to generate re-mastered animation files. The re-mastered animation files are used in the video game to generate gameplay data.

公开(公告)号：WO2022002443A1

公开(公告)日：2022-01-06

申请号：PCT/EP2021/051991

申请日：2021-01-28

Applicant: CATTLE EYE LTD

Inventor： CANNING, Terry ,  ASKEW, Adam ,  MCMILLAN, Ryan ,  THOMPSON, Ian

IPC: G06T7/246 ,  A01K29/00 ,  G06K9/00 ,  A01K29/005 ,  G06T2207/10016 ,  G06T2207/20084 ,  G06T2207/30232 ,  G06V40/103 ,  G06V40/25

Abstract: The present invention provides a method and a system for monitoring the mobility levels of individual farm animals and accordingly determining their corresponding mobility score. The mobility score may be indicative of the health and/or welfare status of the animals. The present invention processes a 2D video recording obtained from an imaging device to detect the movement of individual animals through a space. The video recording is segmented over a set of individual frames and in each frame the individual instances of the animal appearing in the vide frame are detected. The detected instances of each animal over a number of frames are grouped together. From each detected instance of an individual animal a set of reference points are extracted. The reference points are associated with location on the animal body. The present invention determines the mobility score of each animal by monitoring the relative position between reference points in each frame and the relative position of each reference point across the set of individual frames associated with an animal.