Skip to main navigation Skip to search Skip to main content

RGB-D object recognition using the knowledge transferred from relevant RGB images

  • Depeng Gao
  • , Rui Wu*
  • , Jiafeng Liu
  • , Qingcheng Huang
  • , Xianglong Tang
  • , Peng Liu
  • *Corresponding author for this work
  • Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The availability of depth images provides a new possibility to solve the challenging object recognition problem. However, when there is not enough labeled data, we cannot learn a discriminative classifier even using depth information. To solve this problem, we extend LCCRRD method by kernel trick. First, we construct two RGB classifiers with all labeled RGB images from source and target domain. The significant samples for both classifier are boosted and the non-significant ones are inhibited by exploiting the relationship between two domains. In this process, the knowledge of source RGB classifier can be transferred to target RGB classifier effectively. Then to improve the performance of RGB-D classifier by applying the knowledge from source domain, the predicted results of RGB-D classifier are made consistent to target RGB classifier. Furthermore all the parameters are optimized in a unified objective function. Experiments on four cross-domain dataset pairs shows that our approach is indeed effective and promising.

Original languageEnglish
Title of host publicationNeural Information Processing - 24th International Conference, ICONIP 2017, Proceedings
EditorsDerong Liu, Shengli Xie, Dongbin Zhao, Yuanqing Li, El-Sayed M. El-Alfy
PublisherSpringer Verlag
Pages642-651
Number of pages10
ISBN (Print)9783319701356
DOIs
StatePublished - 2017
Event24th International Conference on Neural Information Processing, ICONIP 2017 - Guangzhou, China
Duration: 14 Nov 201718 Nov 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10639 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference24th International Conference on Neural Information Processing, ICONIP 2017
Country/TerritoryChina
CityGuangzhou
Period14/11/1718/11/17

Keywords

  • Depth images
  • RGB-D object recognition
  • Transfer learning

Fingerprint

Dive into the research topics of 'RGB-D object recognition using the knowledge transferred from relevant RGB images'. Together they form a unique fingerprint.

Cite this