Extended CRC: Face Recognition with a Single Training Image per Person via Intraclass Variant Dictionary

Lin, Guanghui; Xie, Mei; Mao, Ling

doi:10.1587/transinf.e96.d.2290

Cited by 4 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the additional dictionary may be able to cover the possibly large changes between the test sample and the corresponding training samples from the same class, ESRC can be applied to single-shot recognition problems where only a single training sample is available for each class. Later on, a non-sparse version Extended CRC (ECRC) [5] was published. The only difference between ECRC and ESRC is that ECRC uses l 2norm instead of l 1 -norm for coefficients regularization.…”

Section: Comparison On Extended Crc Modelsmentioning

confidence: 99%

“…Therefore, for a better understanding and comparison, these two models were both treated as collaborative representation based classification Since the birth of CRC l 2 , more and more attention has been paid to l 2 -norm based regularization for collaborative representation due to its attractive effectiveness and efficiency. Though experiments on both CRC l 2 itself [4] and its extensions [5] have shown their superiority to the sparse representation competitors, there are counterexamples reported as well [4], [6]. The uncertain relative superiority between sparse and non-sparse CRC models confuses people who have limited research experiences on them, and there is still a lack of an in-depth and reliable criterion for preselecting the more promising model for a given task.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Collaborative Representation for Classification, Sparse or Non-sparse?

Wu,

Jarich,

Mukunoki

et al. 2014

Preprint

View full text Add to dashboard Cite

Sparse representation based classification (SRC) has been proved to be a simple, effective and robust solution to face recognition. As it gets popular, doubts on the necessity of enforcing sparsity starts coming up, and primary experimental results showed that simply changing the l 1 -norm based regularization to the computationally much more efficient l 2 -norm based non-sparse version would lead to a similar or even better performance. However, that's not always the case. Given a new classification task, it's still unclear which regularization strategy (i.e., making the coefficients sparse or non-sparse) is a better choice without trying both for comparison. In this paper, we present as far as we know the first study on solving this issue, based on plenty of diverse classification experiments. We propose a scoring function for pre-selecting the regularization strategy using only the dataset size, the feature dimensionality and a discrimination score derived from a given feature representation. Moreover, we show that when dictionary learning is taking into account, non-sparse representation has a more significant superiority to sparse representation. This work is expected to enrich our understanding of sparse/non-sparse collaborative representation for classification and motivate further research activities.

show abstract

Section: Comparison On Extended Crc Modelsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Collaborative Representation for Classification, Sparse or Non-sparse?

Wu,

Jarich,

Mukunoki

et al. 2014

Preprint

View full text Add to dashboard Cite

show abstract

“…ESRC has superior performance to SRC. According to the advantage that the computational complexity of CRC is remarkably lower than that of SRC, enlightened from ESRC, the extended collaborative representation based classification (ECRC) [26] was proposed for undersampled face recognition. ECRC performs better than CRC.…”

Section: Introductionmentioning

confidence: 99%

Extended JSSL for Multi-Feature Face Recognition via Intra-Class Variant Dictionary

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

Auxiliary Dictionary of Diversity Learning for Face Recognition with a Single Sample Per Person

Gan

Huixian

Zeng

et al. 2020

Int. J. Artif. Intell. Tools

View full text Add to dashboard Cite

Face recognition for a single sample per person is challenging due to the lack of sufficient sample information. However, using generic training set to learn an auxiliary dictionary is an effective way to alleviate this problem. Considering generic training sample of diversity, we proposed an algorithm of auxiliary dictionary of diversity learning (ADDL). We first produced virtual face images by mirror images, square block occlusion and grey transform, and then learned an auxiliary dictionary of diversity using a designed objective function. Considering patch-based method can reduce the influence of variations, we seek extended sparse representation with l2-minimization for each probe patch. Experimental results in the CMUPIE, Extended Yale B and LFW datasets demonstrate that ADDL performs better than other related algorithms.

show abstract

Extended CRC: Face Recognition with a Single Training Image per Person via Intraclass Variant Dictionary

Cited by 4 publications

References 9 publications

Collaborative Representation for Classification, Sparse or Non-sparse?

Collaborative Representation for Classification, Sparse or Non-sparse?

Extended JSSL for Multi-Feature Face Recognition via Intra-Class Variant Dictionary

Auxiliary Dictionary of Diversity Learning for Face Recognition with a Single Sample Per Person

Contact Info

Product

Resources

About