Integrating simplified inverse representation and CRC for face recognition

Yingnan Zhao, Xiangjian He, Beijing Chen, Xiaoping Zhao

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


The representation based classification method (RBCM) has attracted much attention in the last decade. RBCM exploits the linear combination of training samples to represent the test sample, which is then classified according to the minimum reconstruction residual. Recently, an interesting concept, Inverse Representation (IR), is proposed. It is the inverse process of conventional RBCMs. IR applies test samples’ information to represent each training sample, and then classifies the training sample as a useful supplement for the final classification. The relative algorithm CIRLRC, integrating IR and linear regression classification (LRC) by score fusing, shows superior classification performance. However, there are two main drawbacks in CIRLRC. First, the IR in CIRLRC is not pure, because the test vector contains some training sample information. The other is the computation inefficiency because CIRLRC should solve C linear equations for classifying the test sample respectively, where C is the number of the classes. Therefore, we present a novel method integrating simplified IR (SIR) and collaborative representation classification (CRC), named SIRCRC, for face recognition. In SIRCRC, only test sample information is fully used in SIR, and CRC is more efficient than LRC in terms of speed, thus, one linear equation system is needed. Extensive experimental results on face databases show that it is very competitive with both CIRLRC and the state-of-the-art RBCM.

Original languageEnglish
Title of host publicationMulti-disciplinary Trends in Artificial Intelligence - 9th International Workshop, MIWAI 2015, Proceedings
EditorsXianghan Zheng, Antonis Bikakis
PublisherSpringer Verlag
Number of pages13
ISBN (Print)9783319261805
Publication statusPublished - 2015
Externally publishedYes
Event9th International Workshop on Multi-disciplinary Trends in Artificial Intelligence, MIWAI 2015 - Fuzhou, China
Duration: 13 Nov 201515 Nov 2015

Publication series

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


Conference9th International Workshop on Multi-disciplinary Trends in Artificial Intelligence, MIWAI 2015


  • Collaborate recognition classification
  • Face recognition
  • Inverse representation

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science (all)


Dive into the research topics of 'Integrating simplified inverse representation and CRC for face recognition'. Together they form a unique fingerprint.

Cite this