Invention Grant
- Patent Title: Method for classification of unique/rare cases by reinforcement learning in neural networks
-
Application No.: US16085989Application Date: 2017-03-15
-
Publication No.: US10776664B2Publication Date: 2020-09-15
- Inventor: Dzmitry Tsishkou , Rémy Bendahan
- Applicant: IMRA EUROPE S.A.S.
- Applicant Address: FR Sophia-Antipolis
- Assignee: IMRA EUROPE S.A.S.
- Current Assignee: IMRA EUROPE S.A.S.
- Current Assignee Address: FR Sophia-Antipolis
- Agency: Kenealy Vaidya LLP
- Priority: com.zzzhc.datahub.patent.etl.us.BibliographicData$PriorityClaim@51845fb1
- International Application: PCT/EP2017/056172 WO 20170315
- International Announcement: WO2017/158058 WO 20170921
- Main IPC: G06K9/00
- IPC: G06K9/00 ; G06K9/62 ; G06N3/04 ; G06N3/08 ; G06N20/20 ; G05D1/00 ; G05D1/02 ; G06K9/46

Abstract:
Some embodiments are directed to a method to reinforce deep neural network learning capacity to classify rare cases, which includes the steps of training a first deep neural network used to classify generic cases of original data into specified labels; localizing discriminative class-specific features within the original data processed through the first deep neural network and mapping the discriminative class-specific features as spatial-probabilistic labels; training a second-deep neural network used to classify rare cases of the original data into the spatial-probabilistic labels; and training a combined deep neural network used to classify both generic and rare cases of the original data into primary combined specified and spatial-probabilistic labels.
Public/Granted literature
- US20190122077A1 METHOD FOR CLASSIFICATION OF UNIQUE/RARE CASES BY REINFORCEMENT LEARNING IN NEURAL NETWORKS Public/Granted day:2019-04-25
Information query