On discrete cosine transform
By Jianqin Zhou. 2011
3,178 citations recorded by the source corpus.
The discrete cosine transform (DCT), introduced by Ahmed, Natarajan and Rao, has been used in many applications of digital signal processing, data compression and information hiding. There are four types of the discrete cosine transform. In simulating the discrete cosine transform, we propose a generalized discrete cosine transform with three parameters, and prove its orthogonality for some new cases. A new type of discrete cosine transform is proposed and its orthogonality is proved. Finally, we propose a generalized discrete W transform with three parameters, and prove its orthogonality for some new cases.
Subfields: Artificial Intelligence, Computer Networks and Communications, Numerical Methods & Scientific Computing, Signal Processing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Dacheng Tao
- Yoshua Bengio
- Percy Liang
- Mohit Bansal
- Quoc V. Le
- Jianfeng Gao
- Ruslan Salakhutdinov
- Sergey Levine
In Computer Networks and Communications
- Francis R. Bach
- Shuchin Aeron
- Pierre Vandergheynst
- Pascal Frossard
- Yehui Tang
- Feng Xue
- José M. Bioucas-Dias
- Akhil Datta-Gupta
In Numerical Methods & Scientific Computing
- Sergey V. Dolgov
- Dmitry V. Savostyanov
- Boris N. Khoromskij
- Yehui Tang
- Feng Xue
- José M. Bioucas-Dias
- Akhil Datta-Gupta
- Abdulbasit K. Al-Talabani
In Signal Processing
- Julien Mairal
- Chin-Hui Lee
- Francis R. Bach
- Johannes Ballé
- Valero Laparra
- Zhen Huang
- Jean Ponce
- Eero P. Simoncelli
Citations in the corpus
- Generating Images with Sparse Representations (2021, methodology): “JPEG (Wallace, 1992) and other popular lossy image compression methods use the discrete cosine transform (DCT) (Ahmed et al., 1974; Rao & Yip, 1990) to separate spatial frequencies of an image, and encode them with controllable resource budgets (e.g., the “quality” parameter).”
- End-to-end optimization of nonlinear transform codes for perceptual quality (2016, background): “KLT for a separable autoregressive process of order 1 [12].”
- Mel Frequency Cepstral Coefficient and its Applications: A Review (2022, methodology): “In the MFCC process, the DCT is applied on the Mel filter bank to select most accelerative coefficients or to separate the relationship in the log spectral magnitudes from the filter-bank [14].”
- Vertex-Frequency Analysis on Graphs (2013, methodology): “…possible choice of associated orthonormal eigenvectors is These graph Laplacian eigenvectors, which are shown in Figure 5(b), are the basis vectors in the Discrete Cosine Transform (DCT-II) transform [17] These eigenvectors correspond to the columns of the Discrete Fourier Transform (DFT) matrix.”
- Compressed Sensing (2014, neutral): “Por ejemplo, para señales de audio puede ser útil una base de tipo Fourier, como la Transformada Discreta Coseno (DCT) [1], y para las imágenes una base de funciones de tipo “Wavelets” [2].”
- Superfast Fourier Transform Using QTT Approximation (2012, methodology): “Example 2 Given a vector x , compute the orthogonal DCT-II (discrete cosine even type-2) transform, defined as follows The history of this and related cosine transforms, including the applications to signal and image processing, can be found in a nice review [53].”
- Estimators of Fractal Dimension: Assessing the Roughness of Time Series and Spatial Data (2011, methodology): “The definition of B̂(ω) is similar in spirit to the so-called type-II discrete cosine transform (DCT-II); see Ahmed, Natarajan and Rao (1974) and Strang (1999) for background.”
- Tensor-Tensor Products with Invertible Linear Transforms (2015, positive): “Some of these transforms, such as the discrete cosine transform, are known to be highly useful in image processing [13].”
- Graph Signal Processing: Overview, Challenges, and Applications (2017, background): “frequencies associated to those regular graphs [184] (see also Section I-B).”
- A generalized grid connectivity–based parameterization for subsurface flow model calibration (2011, methodology): “For the DCT, maximal compression power results from its asymptotic convergence to PCA of a first‐order stationary Markov process [Ahmed et al., 1974].”
- Activity recognition from acceleration data based on discrete consine transform and SVM (2009, methodology): “Discrete cosine transform [7] is an important orthogonal transform and its performance has been found to be asymptotically equivalent to the optimal Karhunent-Loeve transform for signal decorrelation.”
- A General and Adaptive Robust Loss Function (2017, methodology): “For this we use the DCT [1] and the CDF 9/7 wavelet decomposition [7], both with a YUV colorspace.”
- Intensity-Based Image Registration by Minimizing Residual Complexity (2010, methodology): “Consider the first-order derivative regularizer (finite differences matrix) , the corresponding self-adjoint operator , which appears after taking the gradient of , is a second-order derivative matrix (discrete Laplacian), which decomposes as [20]”
- Robust smoothing of gridded data in one and higher dimensions with missing values (2010, neutral): “(8), are given by Γ i,i = 1 + s (2 − 2 cos ((i − 1)π /n)) 2 −1 and Γ i,j = 0 if i = j. (10) It is worth noting that U T and U are actually n-by-n type-2 discrete cosine transform (DCT) and inverse DCT matrices, respectively (Strang, 1999).”
- Automatic Analysis of Facial Actions: A Survey (2019, neutral): “Cosine Transform (DCT) features [1] and Haar-like features [136].”
- Hyperspectral Pansharpening: A Review (2015, background): “Learning the dictionary from the observed images instead of using predefined bases [62]–[64] generally improves image representation [65], which is preferred in most scenarios.”
- Multi-Wiener SURE-LET Deconvolution (2013, positive): “Yet, periodic extensions may generate unwanted artifacts (discontinuity at the boundaries), so symmetric boundary extensions are often preferred in practice [25], [60], and [61].”