Presentation 2011-03-07
Image Recognition Using Linguistic Resource Based on Scene Estimation
Osamu OURA, Masafumi HAGIWARA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose a new method of object recognition using a language resource based on scene estimation. First, object recognition is performed and the scene is estimated using the object recognition. Then the result of the scene recognition is feedbacked to use in the next object recognition. In the first step, object recognition is performed for an input image. In the second step, two-layer neural network is composed. In the third step, feedback is carried out. Confidence score for each scene is calculated for each feedback. In the last step, the most plausible object recognition result is output using the confidence scores. We carried out experiments. It has shown that the object recognition performance is greatly improved by the proposed feedback mechanism.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Image Recognition / Neural Network / Context / Google N-gram
Paper # NC2010-135
Date of Issue

Conference Information
Committee NC
Conference Date 2011/2/28(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Image Recognition Using Linguistic Resource Based on Scene Estimation
Sub Title (in English)
Keyword(1) Image Recognition
Keyword(2) Neural Network
Keyword(3) Context
Keyword(4) Google N-gram
1st Author's Name Osamu OURA
1st Author's Affiliation Faculty of Science and Technology, Keio University()
2nd Author's Name Masafumi HAGIWARA
2nd Author's Affiliation Faculty of Science and Technology, Keio University
Date 2011-03-07
Paper # NC2010-135
Volume (vol) vol.110
Number (no) 461
Page pp.pp.-
#Pages 6
Date of Issue