Invention Grant
- Patent Title: Neural network-based action detection
-
Application No.: US16348420Application Date: 2017-10-31
-
Publication No.: US11003949B2Publication Date: 2021-05-11
- Inventor: Cuiling Lan , Wenjun Zeng , Sijie Song , Junliang Xing
- Applicant: MICROSOFT TECHNOLOGY LICENSING, LLC
- Applicant Address: US WA Redmond
- Assignee: MICROSOFT TECHNOLOGY LICENSING, LLC
- Current Assignee: MICROSOFT TECHNOLOGY LICENSING, LLC
- Current Assignee Address: US WA Redmond
- Agency: Merchant & Gould P.C.
- Priority: CN201610987537.0 20161109
- International Application: PCT/US2017/059139 WO 20171031
- International Announcement: WO2018/089221 WO 20180517
- Main IPC: G06K9/62
- IPC: G06K9/62 ; G06K9/00 ; G06N3/04 ; G06N3/08

Abstract:
Various implementations of the subject matter described herein relate to a neural network-based action detection. There is provided an action detection scheme using a neural network. The action detection scheme can design and optimize the neural network model based on respective importance of different frames such that frames that are more important or discriminative for action recognition tend to be assigned with higher weights and frames that are less important or discriminative for action recognition tend to be assigned with lower weights.
Public/Granted literature
- US20200074227A1 NEURAL NETWORK-BASED ACTION DETECTION Public/Granted day:2020-03-05
Information query