A local spectral distribution approach to face recognition

Pothos, Vasileios K.; Theoharatos, Christos; Economou, George

doi:10.1016/j.cviu.2012.01.006

Cited by 3 publications

(3 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…16,17 This approach is referred to as Gabor decomposition and has been applied in a broad range of applications in the field of computer vision technology. 18,19 In the work by Recher et al, 9 a bifrequency Gabor-based approach was already introduced to locate single-and double-band SHG patterns of manually selected onedimensional sarcomere intensity profiles. Here we propose to use an alternative Gabor approach in which regular, singleband sarcomeres are located by analyzing full images for the frequency related to the average sarcomere length.…”

Section: Introductionmentioning

confidence: 99%

Fully automated muscle quality assessment by Gabor filtering of second harmonic generation images

et al. 2016

View full text Add to dashboard Cite

Although structural changes on the sarcomere level of skeletal muscle are known to occur due to various pathologies, rigorous studies of the reduced sarcomere quality remain scarce. This can possibly be explained by the lack of an objective tool for analyzing and comparing sarcomere images across biological conditions. Recent developments in second harmonic generation (SHG) microscopy and increasing insight into the interpretation of sarcomere SHG intensity profiles have made SHG microscopy a valuable tool to study microstructural properties of sarcomeres. Typically, sarcomere integrity is analyzed by fitting a set of manually selected, one-dimensional SHG intensity profiles with a supramolecular SHG model. To circumvent this tedious manual selection step, we developed a fully automated image analysis procedure to map the sarcomere disorder for the entire image at once. The algorithm relies on a single-frequency wavelet-based Gabor approach and includes a newly developed normalization procedure allowing for unambiguous data interpretation. The method was validated by showing the correlation between the sarcomere disorder, quantified by the M-band size obtained from manually selected profiles, and the normalized Gabor value ranging from 0 to 1 for decreasing disorder. Finally, to elucidate the applicability of our newly developed protocol, Gabor analysis was used to study the effect of experimental autoimmune encephalomyelitis on the sarcomere regularity. We believe that the technique developed in this work holds great promise for high-throughput, unbiased, and automated image analysis to study sarcomere integrity by SHG microscopy.

show abstract

Section: Introductionmentioning

confidence: 99%

Fully automated muscle quality assessment by Gabor filtering of second harmonic generation images

et al. 2016

View full text Add to dashboard Cite

show abstract

“…However, such systems are specialized on single problems, while in real applications, we are required to deal with any possible imperfection, even added together. In this perspective, a big effort has been done yielding FR systems [27,30,40] achieving high performance under uncontrolled conditions. But again, when trying to adopt them in real applications, other problems arise.…”

Section: Introductionmentioning

confidence: 99%

“…On the contrary, we can find works where the training phase requires only one image per subject (facing the so-called small sample size problem) [18,28], but then the performances are too poor. Another question concerns the pre-processing step of face cropping: most approaches [27,30] present results on face images cropped using manual annotated landmarks, but this is not indicative of the performance we would have applying such methods on automatically detected faces. In fact, it has been amply demonstrated that the system performance decreases drastically in the presence of misalignments [6].…”

Section: Introductionmentioning

confidence: 99%

Robust face recognition using sparse representation in LDA space

Adamo

Grossi

Lanzarotti

et al. 2015

Machine Vision and Applications

View full text Add to dashboard Cite

In this article, we address the problem of face recognition under uncontrolled conditions. The proposed solution is a numerical robust algorithm dealing with face images automatically registered and projected via the linear discriminant analysis (LDA) into a holistic low-dimensional feature space. At the heart of this discriminative system, there are suitable nonconvex parametric mappings based on which a fixed-point technique finds the sparse representation of test images allowing their classification. We theoretically argue that the success achieved in sparsity promoting is due to the sequence of values imposed on a characteristic parameter of the used mapping family. Experiments carried out on several databases (ORL, YaleB, BANCA, FRGC v2.0) show the robustness and the ability of the system for classification purpose. In particular, within the area of sparsity promotion, our recognition system shows very good performance with respect to those achieved by the stateof-the-art 1 norm-based sparse representation classifier (SRC), the recently proposed 2 norm-based collaborative representation classifier (CRC), the LASSO-based sparse decomposition technique, and the weighted sparse represen-B Raffaella Lanzarotti tation method (WSRC), which integrates sparsity and data locality structure.

show abstract

Sparse Recovery by Nonconvex Lipshitzian Mappings

Adamo¹

2013

View full text Add to dashboard Cite

A local spectral distribution approach to face recognition

Cited by 3 publications

References 49 publications

Fully automated muscle quality assessment by Gabor filtering of second harmonic generation images

Fully automated muscle quality assessment by Gabor filtering of second harmonic generation images

Robust face recognition using sparse representation in LDA space

Sparse Recovery by Nonconvex Lipshitzian Mappings

Contact Info

Product

Resources

About