Invention Grant
- Patent Title: Fine-grained image classification by exploring bipartite-graph labels
-
Application No.: US15095260Application Date: 2016-04-11
-
Publication No.: US10074041B2Publication Date: 2018-09-11
- Inventor: Feng Zhou , Yuanqing Lin
- Applicant: NEC Laboratories America, Inc.
- Applicant Address: JP
- Assignee: NEC Corporation
- Current Assignee: NEC Corporation
- Current Assignee Address: JP
- Agent Joseph Kolodka
- Main IPC: G06K9/62
- IPC: G06K9/62 ; G06K9/66 ; G06T7/00 ; G06N3/08 ; G06K9/46

Abstract:
Systems and methods are disclosed for deep learning and classifying images of objects by receiving images of objects for training or classification of the objects; producing fine-grained labels of the objects; providing object images to a multi-class convolutional neural network (CNN) having a softmax layer and a final fully connected layer to explicitly model bipartite-graph labels (BGLs); and optimizing the CNN with global back-propagation.
Public/Granted literature
- US20160307072A1 Fine-grained Image Classification by Exploring Bipartite-Graph Labels Public/Granted day:2016-10-20
Information query