Preliminary Studies on a Large Face Database

Yip, Benjamin; Bingham, Garrett; Kempfert, Katherine C.; Fabish, Jonathan; Kling, Troy; Chen, Cuixian; Wang, Yishi

doi:10.48550/arxiv.1811.06446

Cited by 1 publication

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since then, many studies on Morph-II have adopted such an evaluation protocol. Based on discussions in Guo et al [48], a new automatic subsetting scheme is proposed in [71], aiming to automatically ensure independent training and testing sets. Additionally, inconsistencies in age, gender, and race in Morph-II have been identified and corrected in [71].…”

Section: Subsetting Schemementioning

confidence: 99%

“…Based on discussions in Guo et al [48], a new automatic subsetting scheme is proposed in [71], aiming to automatically ensure independent training and testing sets. Additionally, inconsistencies in age, gender, and race in Morph-II have been identified and corrected in [71]. After following the steps to clean MORPH-II outlined in [71], we apply the automatic subsetting scheme, summarized in Figure 4 and described below.…”

Section: Subsetting Schemementioning

confidence: 99%

“…Additionally, inconsistencies in age, gender, and race in Morph-II have been identified and corrected in [71]. After following the steps to clean MORPH-II outlined in [71], we apply the automatic subsetting scheme, summarized in Figure 4 and described below.…”

Section: Subsetting Schemementioning

confidence: 99%

“…Separately within each subset S 1 and S 2 , we fix the ratios of white (W) to black (B) images as 1:1 and male (M) to female (F) images as 3:1. Further, S 1 and S 2 have been selected such that the age distributions within each set are similar (details shown in [71]). The gender and race summaries for the subsetting scheme are shown in Table 2.…”

Section: Subsetting Schemementioning

confidence: 99%

“…Figure 4. Flowchart representing our subsetting scheme[71] for MORPH-II, which improves the one from[48].…”

mentioning

confidence: 99%

See 4 more Smart Citations

A Comparison Study on Nonlinear Dimension Reduction Methods with Kernel Variations: Visualization, Optimization and Classification

Kempfert¹,

Wang²,

Chen³

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

Because of high dimensionality, correlation among covariates, and noise contained in data, dimension reduction (DR) techniques are often employed to the application of machine learning algorithms. Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), and their kernel variants (KPCA, KLDA) are among the most popular DR methods. Recently, Supervised Kernel Principal Component Analysis (SKPCA) has been shown as another successful alternative. In this paper, brief reviews of these popular techniques are presented first. We then conduct a comparative performance study based on three simulated datasets, after which the performance of the techniques are evaluated through application to a pattern recognition problem in face image analysis. The gender classification problem is considered on MORPH-II and FG-NET, two popular longitudinal face aging databases. Several feature extraction methods are used, including biologically-inspired features (BIF), local binary patterns (LBP), histogram of oriented gradients (HOG), and the Active Appearance Model (AAM). After applications of DR methods, a linear support vector machine (SVM) is deployed with gender classification accuracy rates exceeding 95% on MORPH-II, competitive with benchmark results. A parallel computational approach is also proposed, attaining faster processing speeds and similar recognition rates on MORPH-II. Our computational approach can be applied to practical gender classification systems and generalized to other face analysis tasks, such as race classification and age prediction.

show abstract

Section: Subsetting Schemementioning

confidence: 99%