Age Estimation Based on AAM and 2D-DCT Features of Facial Images

Günay, Asuman; Nabiyev, Vasif V.

doi:10.14569/ijacsa.2015.060217

Cited by 8 publications

(7 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to estimate the age accurately, the design of age estimation system requires careful selection of the face feature. To do so, Gunay et al [84] stated some issues that need to take into consideration:…”

Section: Discussionmentioning

confidence: 99%

“…The age was estimated in two steps determine a suitable aging pattern from aging pattern subspace, and then find the specific age position within aging pattern [73]. However, the process of reconstruction of missing face images produces Hybrid Suo et al [80] Guo et al [39] Luu et al [81] Choi et al [40], Guo et al [82] Weng et al [83], Tharwat et al [19] Gunay and Nabiyev [84], Huerta et al [85] Gunay and Nabiyev [86], Bekhouche et al [49], Qiu [87] Ng et al [60], [6] a highly redundant face image, and this can increase the prediction error. In addition, the personalized method for age estimation such as AGES and methods in [77], [74], [71] ccan improve the performance of age estimation.…”

Section: A Features Extractionmentioning

confidence: 99%

“…Gunay and Nabiyev [84], [86] conducted several experiments to estimate the age using different local and global methods to produce hybrid-based age estimation algorithms. In 2015 they proposed an age estimation method called "Global and Local feAture based Age estiMation (GLAAM)" that used AAM to extract global features and 2D-DCT for local features.…”

Section: A Features Extractionmentioning

confidence: 99%

“…PCA was applied to reduce the dimensionality after concatenating the local and global features. Finally, the age was estimated with multiple linear regression [84]. In 2016 Gunay and Nabiyev [86] repeated the experiment using different methods for local features.…”

Section: A Features Extractionmentioning

confidence: 99%

“…This sequence fits the ordinal nature of age label. This is a very popular age learning technique it includes many methods that can be used for regression such as SVR [46], Quadratic Regression [72], Multiple Linear Regression [84], and Multilayer Perceptron (MLP) [110]. Since 2007, Geng et al [73] has achieved MAE of 6.77 years on FG-NET using facial appearance features.…”

Section: ) Classificationmentioning

confidence: 99%

See 4 more Smart Citations

Computational Intelligence in Automatic Face Age Estimation: A Survey

Osman

Yap

2019

IEEE Trans. Emerg. Top. Comput. Intell.

View full text Add to dashboard Cite

With the rapid growth of computational intelligence techniques, automatic face age estimation has achieved good accuracy that benefited real-world applications such as access control and monitoring, soft biometrics and information retrieval. Over the past decade, many new algorithms were developed and previous surveys on face age estimation were either outdated or incomplete. Considering the importance of the expanding research in this topic, we aim to provide an up-to-date survey on the face age estimation techniques. First we summarise the stateof-the-art databases and the performance metrics for face age estimation. Then we review the age estimation techniques based on three categories of face features (local, global and hybrid) and discuss different types of age learning algorithms. Finally, we identify the challenges and provide new insights for future research directions of fully automated face age estimation.

show abstract

Section: Discussionmentioning

confidence: 99%