Face Recognition Using the Discrete Cosine Transform

By Ziad M. Hafed, Martin David Levine. International Journal of Computer Vision, 2001

529 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/ea79ae7f32b3d63f1253757aae91149a6e1aa440

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Xilin Chen
2. Shiguang Shan
3. Xiaogang Wang
4. Luc Van Gool
5. Dacheng Tao
6. Wen Gao
7. Yi Yang
8. Wanli Ouyang

Citations in the corpus

• Fusing Local Patterns of Gabor Magnitude and Phase for Face Recognition (2010, background): “Later on, transformation features in the frequency domain or wavelet domain attracted much attention for face recognition, such as DFT [17], discrete cosine transform (DCT) [ 9 ], discrete wavelet transform (DWT) [5], and Gabor wavelet [6], [13].”
• Illumination compensation and normalization for robust face recognition using discrete cosine transform in logarithm domain (2006, methodology): “It should be noted that our method is different from the DCT applied for dimensionality reduction in face recognition [20], in which only low-frequency coefficients are used as facial features.”
• Faster Neural Networks Straight from JPEG (2018, background): “Several works have employed spatial frequency decomposition and other compressed representations for image processing without using deep learning [22, 18, 8, 5, 7].”
• Hierarchical Ensemble of Global and Local Classifiers for Face Recognition (2009, result): “The conflicting results of the two classifiers interestingly reflect the above-mentioned different roles of global and local information, which suggests that ideal classifier should be the combination of the two “experts.””
• DCT-Based Iris Recognition (2007, methodology): “In particular, the DCT has been shown to produce good results on face recognition [27], [28], where it has been used as a less computationally intensive replacement for the Karhunen-Loeve transform (KLT), which is an optimal technique according to the least squares metric for projecting a large amou”