公开(公告)号：US20230046066A1

公开(公告)日：2023-02-16

申请号：US17960760

申请日：2022-10-05

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Adrian BULAT ,  Georgios TZIMIROPOULOS ,  Juan-Manuel PEREZ-RUA ,  Swathikiran SUDHAKARAN ,  Brais MARTINEZ

IPC: G06V20/40 ,  G06V10/77

Abstract: Broadly speaking, the present techniques generally relate to a method and apparatus for video recognition, and in particular relate to a computer-implemented method for performing video recognition using a transformer-based machine learning, ML, model. Put another way, the present techniques provide new methods of image processing in order to automatically extract feature information from a video.

公开(公告)号：US20230145150A1

公开(公告)日：2023-05-11

申请号：US17421073

申请日：2021-02-16

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Brais MARTINEZ ,  Tao XIANG ,  Victor Augusto ESCORCIA ,  Juan PEREZ-RUA ,  Xiatian ZHU ,  Antoine TOISOUL

IPC: G06V40/20 ,  G06V10/82 ,  G06V10/77 ,  G06V20/40

CPC classification number: G06V40/20 ,  G06V10/82 ,  G06V10/77 ,  G06V20/40

Abstract: Broadly speaking, the present techniques relate to a method and apparatus for performing action recognition, and in particular to a computer-implemented method for performing action recognition on resource-constrained or lightweight devices such as smartphones. The ML model may be adjusted to achieve required accuracy and efficiency levels, while also taking into account the computational capability of the apparatus that is being used to implement the ML model. One way is to adjust the number of channels assigned to the first set of channels, i.e. the full temporal resolution channels. Another way is to adjust the point in the ML model where the temporal pooling layer or layers are applied.