公开(公告)号：US20190147284A1

公开(公告)日：2019-05-16

申请号：US16189974

申请日：2018-11-13

Applicant: QUALCOMM Technologies, Inc.

Inventor： Kirill GAVRILYUK ,  Amir GHODRATI ,  Zhenyang LI ,  Cornelis Gerardus Maria SNOEK

IPC: G06K9/32 ,  G06K9/00 ,  G06F17/30 ,  G06T7/73

Abstract: A method of pixel-wise localization of an actor and an action in a sequence of frames includes receiving a natural language query describing the action and the actor. The method also includes receiving the sequence of frames. The method further includes localizing the action and the actor in the sequence of frames based on the natural language query.

公开(公告)号：US20220156514A1

公开(公告)日：2022-05-19

申请号：US17454743

申请日：2021-11-12

Applicant: QUALCOMM Technologies, Inc.

Inventor： Kirill GAVRILYUK ,  Mihir JAIN ,  Cornelis Gerardus Maria SNOEK

IPC: G06K9/62 ,  G06K9/46 ,  G06K9/00 ,  G06N3/08

Abstract: Certain aspects of the present disclosure provide techniques for training a first model based on a first labeled video dataset; generating a plurality of action-words based on output generated by the first model processing motion data in videos of an unlabeled video dataset; defining labels for the videos in the unlabeled video dataset based on the generated action-words; and training a second model based on the labels for the videos in the unlabeled video dataset.