Loading...

Signal extrapolation for image and video error concealment using gaussian processes with adaptive nonstationary kernels

Asheri, H ; Sharif University of Technology | 2012

984 Viewed
  1. Type of Document: Article
  2. DOI: 10.1109/LSP.2012.2213593
  3. Publisher: IEEE , 2012
  4. Abstract:
  5. In this letter, a new adaptive Gaussian process (GP) frame work for signal extrapolation is proposed. Signal extrapolation is an essential task in many applications such as concealment of corrupted data in image and video communications. While possessing many interesting properties, Gaussian process priors with inappropriate stationary kernels may create extremely blurred edges in concealed areas of the image. To address this problem, we propose adaptive non-stationary kernels in a Gaussian process framework. The proposed adaptive kernel functions are defined based on the hypothesized edges of the missing areas. Experimental results verify the effectiveness of the proposed method compared to the existing state of the art algorithms, based on objective and subjective evaluations
  6. Keywords:
  7. Adaptive nonstationary kernel ; Bayesian inference ; Gaussian process ; Multidirectional extrapolation ; Adaptive kernels ; Bayesian inference ; Blurred edges ; Corrupted data ; Frame-work ; Gaussian processes ; Gaussian process priors ; Image and video communication ; Multi-directional ; Nonstationary ; Signal extrapolation ; State-of-the-art algorithms ; Subjective evaluations ; Video error concealment ; Bayesian networks ; Extrapolation ; Gaussian noise (electronic) ; Inference engines ; Gaussian distribution
  8. Source: IEEE Signal Processing Letters ; Volume 19, Issue 10 , 2012 , Pages 700-703 ; 10709908 (ISSN)
  9. URL: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?reload=true&arnumber=6269922